UNRISD
UNITED NATIONS RESEARCH INSTITUTE FOR SOCIAL DEVELOPMENT
Welfare and Democracy in Latin America: The
Development, Crises and Aftermath of Universal, Dual
and Exclusionary Social States
Fernando Filgueira
Acadaemic Coordinator, IPES, Universidad Católica del Uruguay
May 2005
prepared for the UNRISD Project on Social Policy and Democratization
DRAFT WORKING DOCUMENT
Do not cite without the authors’ approval
The United Nations Research Institute for Social Development (UNRISD) is an autonomous agency
engaging in multidisciplinary research on the social dimensions of contemporary problems affecting
development. Its work is guided by the conviction that, for effective development policies to be
formulated, an understanding of the social and political context is crucial. The Institute attempts to provide
governments, development agencies, grassroots organizations and scholars with a better understanding
of how development policies and processes of economic, social and environmental change affect
different social groups. Working through an extensive network of national research centres, UNRISD aims
to promote original research and strengthen research capacity in developing countries.
Current research programmes include: Civil Society and Social Movements; Democracy, Governance and
Human Rights; Identities, Conflict and Cohesion; Social Policy and Development; and Technology,
Business and Society.
A list of the Institute’s free and priced publications can be obtained by contacting the Reference Centre.
UNRISD, Palais des Nations
1211 Geneva 10, Switzerland
Tel: (41 22) 9173020
Fax: (41 22) 9170650
E-mail: info@unrisd.org
Web: http://www.unrisd.org
Copyright © United Nations Research Institute for Social Development (UNRISD).
This is not a formal UNRISD publication. The responsibility for opinions expressed in signed studies rests
solely with their author(s), and availability on the UNRISD Web site (http://www.unrisd.org) does not
constitute an endorsement by UNRISD of the opinions expressed in them. No publication or distribution of
these papers is permitted without the prior authorization of the author(s), except for personal use.
2
I wish to thank the insightful comments of Yusuf Bangura, Jorge Papadópulos and Federico
Tobar on an earlier version of this draft. I also wish to aknowledge, beyond the appropriate
references and quotes, the inspiration, analyses and data that came from the studies of Carmelo
Mesa Lago and Evelyne Huber. Errors, remain, of course, my responsibility.
3
Introduction
Citizenship in Latin America, be it civil, political or social has been elusive and fragile
throughout its 20th century history. Coerced and semi coerced labor along racial lines in the late 19th and
early 20th century, extreme levels of inequality, a centralist and patrimonialistic institutional and cultural
tradition and elites that regarded themselves as subjects of privilege rather than citizens with equal rights
and duties, conspired to create states that rarely considered the fate of the people. Only when the
oligarchic states broke down and gave way to the need for a modern bureaucracy and a modern state and
then to a new development model and to something approximating mass politics (not necessarily neither
usually mass democracies) did a social dimension of the state emerge. Among the three processes
mentioned –the development of modern states and bureaucracies, the emergence of a new development
model and mass politics with democratic interludes- the first two processes were, in this early stage,
definitely more important in defining and shaping the social states of Latin America than the latter,
especially if we require of the latter the connotation not merely of mass politics but also a resemblance to
democratic mass politics.
Indeed, the history of the Latin American developmental social state is the history of Statecraft and of the
Import Substitution Model and its political manifestations, mostly authoritarian and corporatist, but rarely
democratic. It is more the history of elite accommodation, elite’s state building and elite’s attempts to coopt and control non-elite sectors than a history of popular achievements and shaping from below. This
does not mean that social and labor organizations, parties, and popular struggles played no role in the
building and especially in the expansion of the region social states, but rarely were popular based political
and social actors seated at the driver’s seat. They were, yes, in some cases central coalitional partners of
the populist regimes of the 30s, 40s and 50s, but under diverse forms of cooptation, clientelism and
patrimonialism, and they entered the system rarely as a unified grass roots movement, but rather as
fragmented narrow lobbying forces.
In this paper I will first go over some basic quantitative data with the single purpose of suggesting the
relevance of the democracy-welfare hypothesis regarding social policy effort. Secondly I delve in depth
into the development of the Social States, linking such development to the characteristics of their Import
Substitution models and the politics of specific countries. I propose a typology that defines three different
groups of social states in Latin America and try to unveil de political and developmental determinants of
their emergence and expansion until the 1970s, when the ISI model broke down and the political regimes
of the most advanced social state changed drastically. Finally I undertake the analyses of the social state
transformation in the last two decades and a half, looking at the era of neglect in the eighties and the
4
liberal turn of the nineties with an emphasis o social security reform. In that section I attempt to show
how democratic policies had both a positive effect (moderating the liberal character of reform) and a
negative effect (allowing for the persistence of privileged groups within the system).
SOME SUGGESTIVE DATA ON SOCIAL POLICY EFFORTS AND DEMOCRACY
To claim that democracy and democratic struggles have been, empirically, a rare building block
of the region’s welfare regimes, does not mean in any way that democracy or the lack of it is irrelevant or
has been irrelevant regarding social states. Both country specific analyses and cross-national evidence
suggest that both the formal rules of democracy and the quality and level of political incorporation that
the Latin-American political regimes achieved have played and may even more play in the future a most
important role. I will consider the country specific models of developments and their politics further
ahead, but before, it is worth to delve into some very basic evidence regarding democracy and social state
effort at the cross-national level with the support of some simple quantitative data. When we consider the
effect of stable democratic rules (that is years of uninterrupted democratic rule) between the 1950s and
1990s on social state size, considered as social expenditure over GDP, we can see that there is no relation
whatsoever.
GRAPH 1
GRAPH 2
20
20
PANAMA
PANAMA
BRASIL
ARGENTINA
18
BRASIL
ARGENTINA
18
URUGUAY
URUGUAY
COSTA RICA
16
COSTA RICA
16
14
14
CHILE
CHILE
12
Gasto social /PBI (1990-1991)
12
10
Gasto Social/PBI 1990
VENEZUELA
COLOMBIA
HONDURAS
8
MEXICO
6
4
PARAGUAY
PERUGUATEMALA
Rsq = 0.0891
2
-10
0
10
20
30
40
10
VENEZUELA
8
MEXICO
6
4
GUATEMALA
PERU
PARAGUAY
Rsq = 0.3517
2
2000
50
COLOMBIA
HONDURAS
4000
6000
8000
PBI per cápita, 2000 (PPA en dólares)
Años de democracia ininterrumpida entre 1950 y 1990
5
10000
12000
14000
Source: Author on the basis of data from Przeworski et all, 2000, UNDP Human Development Reports, and CEPAL,
Panorama Social de América Latina, 2000, 2001, 2002.
This evidence would strengthen the argument that claims that social state’s basic engine is not politics but
economics. Indeed the relationship between GDP and percentage of GDP destined to social spending,
presents a far more robust relationship than that of democracy and social spending as a percentage GDP.
Of course, GDP per capita alone leaves much unexplained. Costa Rica and Mexico are two cases were it
is evident how little GDP helps us explain social spending. Yet when we consider, not social expenditure
as a percentage of GDP, but the unstandardized residual of such social spending as a percentage of GDP
on GDP per capita, the hypothesis of the effect of democracy in general and of stable democracy in
particular seems, at least to barely hold its ground.
GRAPH 3
GRAPH 4
10
10
PANAMA
PANAMA
8
8
BRASIL
6
Unstardized residual of social spending as a % of GDP and GDP, 2000
Unstndardized residual of Social spending as a % of GDP and GDP, 2000
BRASIL
6
URUGUAY
COSTA RICA
4
CHILE
COLOMBIA
2
0
ARGENTINA
HONDURAS
-2
PARAGUAY
-4
PERU
VENEZUELA
GUATEMALA
-6
MEXICO EL SALVADO
URUGUAY
COSTA RICA
4
CHILE
COLOMBIA
2
0
ARGENTINA
HONDURAS
-2
PARAGUAY
-4
VENEZUELA
PERU
GUATEMALA
-6
MEXICO EL SALVADO
Rsq = 0.1730
-8
-8
-10
0
10
20
30
40
50
Rsq = 0.1769
-10
Years of democracy 1950-1990
0
10
20
30
40
50
Years of Uninterrupted Democracy, 1950-1990
Source: Author on the basis of data from Przeworski et all, 2000, UNDP Human Development Reports, and
CEPAL, Panorama Social de América Latina, 2000, 2001, 2002.
Yet this is indeed not a very strong relationship, but only one that suggests that democracy and
social policy effort are not altogether unrelated. The finding that really contributes to take the democracywelfare hypothesis seriously in Latin America, is a peculiar one, and one that surprisingly would reclaim
a major role for democracy between the 1950s and 1970s, a period when, while corporatist and
authoritarian rule were extended in the region, democracy of some sort had taken hold an a number of
countries. Where democracy took hold and even more where it survived for relatively large periods,
6
social spending 30 years later clearly goes beyond what would be expected given its GDP. In contrast
were democracy was weak or non-existent between the 50s and 70s, social spending under performs,
again 30 years later, what should be expected given its GDP.
GRAPH 5
Relation between years of uninterrupted democracy between 1950 and 1970 and the unstandardized
residual of social spending over GDP and GDP per capita, 2000
10
PANAMA
Unstandardized residual of Social Spending as a % of GDP and GDP 2000
8
BRASIL
6
URUGUAY
COSTA RICA
4
COLOMBIA
CHILE
2
0
ARGENTINA
HONDURAS
-2
PARAGUAY
-4
PERU
VENEZUELA
GUATEMALA
-6
ELMEXICO
SALVADO
Rsq = 0.5840
-8
-10
0
10
20
30
Years of uninterrupted democracy 1950-1970
Source: Author on the basis of data from Przeworski et all, 2000, UNDP Human Development Reports, and
CEPAL, Panorama Social de América Latina, 2000, 2001, 2002.
Such finding is no proof of a relationship between democracy and social state development. Both
the data and the number cases, not to mention the number of fallacies that any conclusion implies given
the chronological miss-match between the data, render any strong assertion unwarranted. Yet this is
indeed a curious pattern, and no easy alternative explanation seems obvious. The hypothesis that best fits
the data pattern here, is that the relationship between democracy and welfare expansion is strongly path
dependent. When major developmental surges under a given developmental model (the ISI model) were
combined with stable or relatively stable democratic rule, a social policy effort above and beyond the
7
mean expected social effort given a country GDP should be expected to emerge and to be sustained
despite regime changes in the future. If this hypothesis is plausible, then the present day democratic
hegemony in the region, might well be the basis for a new push regarding social policy efforts. As I will
also discuss later it might also help, not merely to increase social spending, but also provide a new
architecture to the Latin-American Social State. One that is less stratified and more universalistic.
The future of the social state in Latin America will, hopefully be, the history of its future
democratic struggles. Never as before, has the region known such an hegemony of democracy and such
an expansion of basic political incorporation measured in terms of clean, regular and free elections were
all or most of the adult population can and does vote. In this single fact rests, maybe, the biggest hope for
social citizenship in Latin America. But as the last 20 to 30 years of social policy crises and reforms have
shown (1970s-1990s), such future might well be impaired, by the constraints posed by global capitalism,
the beliefs and interests of its financial gatekeepers, the perception of threat felt by domestic elites, the
defense of patrimonial privileges of the middle classes, and by the frailty of Latin America’s tax systems
and their political basis of support.
Still as I will try to show there is room for moderate optimism. Latin American social states stand
at the eve of a major transformation. I will claim that after almost two decades of the wrong medicine for
the old social states, political realities and technical possibilities are coming together and might shape a
renewed and better road for the creation of true welfare regimes. There is, of course a distinct more
pessimistic outcome: the increasing disappearance of the state not just as provider of social protections,
but its disappearance as a basic provider of public goods in general (security, basic services, justice
systems, public infrastructure, etc), thus undermining the very notion of nation-state. Structural conditions
and political actors both national and transnational will be key in defining the future routes of the Latin
American States. For the first optimistic road to triumph, it is necessary that elite based and popular based
political parties reach a historical compromise. Yet as I will try to show, such compromise can only be
reached if previously both elite parties and popular based parties are capable of redefining their leadership
regarding their old and new constituencies. Indeed one of the major problems facing the political systems
of Latin America is that parties leaning left and parties leaning right tend to privilege the representation of
the constituencies from the old developmental model, rather than the new real constituencies that emerged
from the breakdown of the Import Substitution Model and the emergence of the new export oriented
model.
8
THE ISI MODEL OF DEVELOPMENT, POLITICAL REGIMES AND ACTORS: THREE
TYPES OF SOCIAL STATES
1. Historical roots and divergent paths.
Latin America does not have a welfare state. This assertion is true on two counts. First, no social
policy system in Latin America can truly be considered a welfare state, secondly, if, we would consider
their social policy systems as something resembling a welfare state1, we would have to speak about at
least three different types of welfare states. Since I choose not to label them as welfare states I will refer
to the system of social policies and social protection as Social States. Latin America, has, indeed, social
states, if by such we understand those aspects and actions that the state develops in the form of a group of
policies concerned with transfers, subsidies and services that resemble a social safety net and a set of
redistributive transfers and services.
Such social state was mostly inexistent by the turn of the 20th century and well into its first twenty
years. What did exist were the embryos of the future social state in the pioneer countries in the form of
insurance schemes for state employees, public health concerns that translated into some sanitary measures
and primary education expansion as a means to discipline and homogenize the rural migrants and the
European immigration. As a region we can only properly refer to a social state starting roughly in the
1930s and then maturing and expanding throughout the 20th century. The four key sectors or areas that
express the beginnings and expansion of the Latin American Social State and its different shapes and
models were and still are: education, health-care, pensions and transfers, and price controls and subsidies.
When considering the evolution of the social states until the 70s I will consider education, social
security and health care. For the analyses of recent transformation, between the 1970s and 1990s, I will
focus on pension reform with lateral comments to health care reform. The mayor issues to be considered
when analyzing the evolution of the social state are expenditure, coverage, stratification and quality of
social services. On the causal forces driving this evolution the critical factors are: the depth and shape of
the ISI model, the political regimes and actors administering such models and the organizational and
political characteristics assumed by the non-elite sectors –middle classes, formal working classes, urban
informal workers and rural workers and small propertied classes.
1
As we shall see later Uruguay and Argentina in the 1960s and early 1970s, and Costa Rica more recently might be
the only countries were it might be appropriate to speak of welfare regimes.
9
I have argued before -and will attempt to build upon my earlier typology- that three distinct forms
of social state developed in Latin America until the 1970s: stratified universalistic, dual and exclusionary.
I also sketched a structural and political explanation that directly linked these types to how economically
inclusionary were their ISI models and to the organization and response elite’s gave to the political
pressures of subordinate sectors. In this paper I shall attempt to go deeper into the political basis of these
models, especially the universal stratified and the dual regimes, and I will also consider their crises and
further evolution. Indeed I will try to show how after the 1970s these three models went into a critical
phase and into a critical juncture giving birth in the early nineties to two new possible models, one
already tested leaning on the neo-liberal creed, the other still struggling to take shape: an egalitarian
exclusionary basic protection social state, or if we are to take risks, what could be labeled an embryonic
Social Democratic Latin-American State. The role of democracy and politics on social expenditure and
on social security reform after the crises of the old models will be the basis for evaluating the prospects
and future of both democracy and the social state in Latin America.
2. A Typology of Social Policy Systems in Latin America until the 1970s2
A particular form of “Keynesian” politics informed a big part of the development models in Latin
America between the 1930s and the 1970s. Theoretically, the Import Substitution Model was articulated
into the “Cepal” paradigm and owed its basic tenets to Raul Prebisch. In such a model, the State assumed
a central role in the process of economic and social development. By extracting taxes from the export of
staples the import of manufactured and capital goods, large property holdings and some consumption
taxes, the state apparatus of the region protected and financed the growth of the industries oriented to the
domestic production through subsidies and diverse protectionist means. In addition, the State had the role
of absorbing the “redundant” work force and of providing the capital for basic works of economic and
social infrastructure. This model allowed an important process of social and economic modernization.
However, it did so with systematic bias in the distribution of benefits of this development; specifically in
welfare policies. Such policies presented a limited development that was strongly biased to the urban
sectors and with preferential coverage for those incorporated into the formal labor market. The rural
workers and those employed in informal and secondary markets were excluded from protection. In
addition, the sectors actually protected clearly showed a marked stratification regarding access, coverage
and quality of the benefits (Mesa-Lago, 1991). State and key services workers had early access to
2
This section draws largely on a previous work of mine, but elaborates further on the political economy of social
state building and considers two new cases that usually appear as outliers in this and other typologies of social
states: Costa Rica and Peru.
10
programs of comprehensive coverage, while “blue collar” workers received protection both later and of
lower quality.
Besides this very rough outline, it is necessary to point out enormous variations in social state
have always existed in Latin America. The development of social policies must be understood in the
frame of the peculiar economic and political development of the region and of the specific nations. Thus,
diverse degrees of expansion of the Import Substitution Model, different export economies, and the
political power of capital, labor, and the middle classes must be considered (Huber, 1995).
Which elements or dimensions promise to offer substantive guidelines of classification of the
social policies systems in the region? The first challenge that every social policy scholar faces in Latin
America is the enormous variety of models and degrees of development of the welfare systems in the
region. This variety is evident when we look at the fiscal efforts allocated to the social services, the
population coverage, the variety of services and transfers, its quality and the distribution of social
expenditure among sectors.
Carmelo Mesa-Lago (1991) offered a first approximation to this problem, introducing a simple
key of classification; its maturity. It allows him to differentiate between pioneer countries, intermediate
countries, and latecomers.
The pioneer countries are characterized today by having a universal, or almost universal,
coverage in basic social protection services, a marked stratification in the quality and conditions of access
to them and a situation of creeping structural deficit in the matter of resources and expenditures.
On the other extreme, the latecomer countries present a low population coverage, a very limited
menu of options in services, and a disproportionate quality and quantity of benefits given to the small
protected sectors.
Within the intermediate countries there are, in his view, two different types that can be
distinguished. Those countries that had advanced toward a mature stratified model, can be distinguished
from those - especially Costa Rica, and to a lesser extent one could argue Panama - that have advanced
forward in levels of coverage and supply and quality of services without falling into the errors of
stratification of the pioneer systems.
The Mesa-Lago’s classification is simple and tries to situate the countries in a continuous
development of the social policies rather than in a typological classification of the social states. This kind
of classification of the welfare systems, which relies heavily on indicators of coverage and expenditure,
has been increasingly criticized based on the literature oriented to the analysis of the Welfare States in
advanced industrial countries (Tittmus, 1958, Korpi,1993, Esping Andersen, 1990, 1994). According to
these authors, a central problem of the analysis of the welfare states is not how much is spent but how it is
11
spent. The role of welfare states in “decommodification”, redistribution and in the shaping of labor
markets, promises, in this perspective, to produce hypothesis and substantive finding that the research
concentrated on the levels of expenditure and coverage has not produced.
This new literature has been undoubtedly successful and, to a great extent, redefined the academic
debate for the industrialized countries. Nevertheless, there is a warning to be made about the application
of these models of analysis to the Latin-American realities. The industrialized countries present universal
coverage and in all cases important levels of expenditure - even though there are important variations. In
Latin America the differences are still far more significant and should remain a basic classifying criteria.
An analyst cannot concentrate on the “how” is spent as a typological guide and ignore the “how much” is
spent, when some countries present expenditures that come up to 18% of the GDP while others barely
reach 8%. Likewise, when the analyst faces population coverage that vary in a rank from 20% to 90% of
the population, she should seriously look into this point, incorporate it as a discriminate criteria and try to
explain it.
Now, this does not imply that the Esping-Andersen’s critique has no application to the LatinAmerican realities and to the academic production surrounding it. On the contrary, the methodological
and theoretical warning that the author imparts has full validity.
In order to understand and compare welfare regimes, a typological classification is required rather
than one of continuum. Likewise, the explanation for differential national development will not be of
lineal and evolutionary character, but of a “path dependent” and structural type. In other words, the
explanation of the nature of the welfare systems in the region has its main keys in structural variables and
in historical sequences rather than in singular temporal starting points and its correspondent degrees of
maturity. Thus, although the central indicators for the building of typologies are, in the region, coverage,
expenditure and, less so, the “how is spent”, does not imply that we have to give up the attempt to
understand the “types” of social states rather than understand “levels” of social states development. For
this reason, the conceptualization and historical explanation of their development will be a typological
one and will emphasize different variables than the maturity reached by the different systems.
The fundamental indicators to be considered, then, will be: coverage (i.e., to some extent the
“how it is spent”), the expenditure (i.e. the “how much is spent”) and its sector distribution and levels of
services given (i.e. an additional approximation of “how it is spent”). We therefore analyze inclusiveness
as well as expenditure. Our interest is on “how” as an issue of sector distribution and population coverage
and, on a second level, as it relates to decommodification.
12
These indicators will be presented together with dimensions concerning social development
levels. These levels, although dependent only partly on transfers and social services, allow us to have a
broader image of the societies in question.
The second challenge is to link these patterns of coverage to de development models (shapes and
depths of the ISI model), to their political regimes and to the political actors and dynamics (both elite and
non-elite) that were present between the 1930s and 1970s in the different countries of Latin America.
i. Stratified Universalism based on contending elites seeking popular support (Uruguay, Argentina,
Chile3)
According to Mesa-Lago’s definition, these three countries are included in the pioneer countries
category. According to his criterion, Brazil should also be included in this group but, as I am going to
argue in the following pages, this country actually belongs until the 1970s, to another category.
The first central dimension that characterized this group is the fact that, by about 1970 all of them
protected (in one way or another) the vast majority of the population by means of a social security system
and basic health services; they had also achieved universal access to primary education and broadened
access to early secondary education to more than half of the population and to upper high school to a
large part of the population. In other words, all these countries offered extended levels of
decommodification in what concerns both the provision of services out of the market and the provision of
monetary benefits to the elderly and those unable to work.
TABLE 1
Selected variables on social policy effort for countries of stratified universalism, circa 1970
% covered by % covered by Social
Social
Security
EAP
3
Social
%
of
children
expenditure under one year
%
enrolled
primary
in % enrolled in
secondary
vaccinated
education of age education of age
total
against
group
group
population
tuberculosis
in Security
in as % GDP,
Argentina 78.9
68.0
16.2
93
105
44
Uruguay
95.4
68.5
17.0
97
112
59
Chile
75.6
67.3
13.7
98
107
39
The Costa Rican case is treated below as a particular for of Universalism.
13
Sources: Mesa-Lago, 1991; World Bank-World Development Report-, 1994, Miranda, 1994, Cominetti,
1994.
The other central characteristic of these systems is the strong stratification of benefits, access
conditions and ranks of protection concerning social security and, to a similar extent, health. State
workers, professionals, urban services workers and urban manufacturing workers accessed protection and
benefits in this order. The quality and access was also stratified following this order. The self-employed,
workers of the informal sector, longstanding unemployed and rural workers had a later access to benefits.
In other words, the decommodification degrees in these countries are directly stratified. However, it is
important to point out that the stratification of social services moderated rather than increased pure labor
market stratification. In simpler terms, if we calculate Gini’s coefficient for primary income/wage and
compare it with the coefficient for income/wage after taxes and transfers, we will find out that the last one
will be more equal than the first one (CEPAL, 2002).
There is another important point to be made. The development of the different systems of social
protection had an important impact over variables relating to social welfare. These advances cannot be
simply attributed to the level of economic development. Indeed while Brazil and Mexico were in the
seventies poorer than Argentina or Uruguay, their levels of social development were far worse than those
of the countries with stratified universalism.
TABLE 2
Selected social development indicators for countries of stratified universalism, circa 1970
%
Households
below
of %
of Illiteracy rate
Households
the below
Infant
Life
Mortality
Expectancy at
the
birth
Poverty line
indigence line
Argentina
8
1
7.4
41
68.4
Uruguay
10
4
10.2
47
68.6
Chile
17
6
11.0
62
64.2
(a) Only urban areas
14
Sources: ECLAC -Panorama Social de América Latina- 1995; ECLAC, Anuario Estadístico de América
Latina- 1980.
The explanation of the peculiar development of these social systems should be sought, not
especially in the countries early development of their social policies, but more emphatically in the
political economy of their national development and in its political administration. In Argentina and
Uruguay, the Import Substitution Model was based on export of primary goods, especially cattle, which
had a low labor force demand and a high international profitability. Both countries were the first ones to
experiment not just mass politics, but also mass democracy. Indeed between 1912 and 1930 in the case of
Argentina and 1917 and 1931 in the case of Uruguay, the radical party and the Colorado party
respectively mounted a vigorous push for social and political modernization.
Regarding the relationship between regimes and social state expansion it is worth noting that
early social state formation is highly correlated with early democratic experiments. The first countries to
develop mass systems of basic education and some attempts at public health as well as the first social
security schemes are Argentina and Uruguay, both countries that knew a democratic experiment as early
as the 1910s and 1920s. Indeed the Yrigoyenista and Batllista years of the 1920s would see not only an
important expansion of vote and an increasing mobilization of middle and working classes, but also the
first clear cut pro labor bills, and the expansion of public education, already started in the late 19th
century. Uruguay and Argentina both benefited from an export economy that rested on a low intensity
labor demand, making the tolerance of elites to both democracy and urban social protection quite high.
The case of Chile is quite complex. A pioneer, as Argentina and Uruguay were, Chile had to build its
social state on much tougher structural basis than Argentina and Uruguay. It may seem a paradox, that in
Chile it was Ibañez, a military leader through a coup détat, the one that would be responsible for setting
the basis of Chile’s social state. While Alessandri in the 1920s would be the first one to push for social
security schemes, public works, labor laws and the expansion of public health and education, it would
finally be his ally in the military that would see his proposal through after the coup of 1927 until 1931.
Yet the Chilean military were far more that a simple expression of the oligarchy as it had happened in
other countries. They were a major player of a previously limited but nonetheless relevant party system
that had through parliamentary debates and struggles shaped the way to modern social legislation.
Even though Argentina’s experiment with democracy closed in the thirties, only to become a
stable democracy again in the eighties Argentina was already and never seized to be a nation of mass
politics. Despite almost 15 odd years of authoritarianism after the Uriburu coup d’etat of 1930 and the
Justo governments the trade unions remained important actors and due to the industrialization process
15
grew even more. It was in the early 1940s that strong and ultimately successful attempts at co-opting the
trade union movement were made from the government. The ministry of labor of the military regime that
started in 1943, Peron, would be the creator of a corporatist alliance between the workers, a new
movement/party and the state. Indeed Peronismo between 1946 and 1955 would capitalize and definitely
co-opt what were already important and powerful trade unions. With this alliance and the return to
electoral mass politics (the jury is still out on the democratic nature of Peronismo) Peron and its
movement would reform the labor code, grant family allowances, establish a minimum wage, provide
accident insurance and restrict the dismissal of workers, as well as later on finance the health programs
known as Obras Sociales, expand markedly public health and expand social security. At the same time
that this golden age of social and labor legislation took hold the autonomy of trade unions diminished.
The single most important effect of the authoritarian period of the thirties and early forties in Argentina
would not be to end the process of social state expansion, even though it did bring a parenthesis to it. It
would be to shift what seemed during Yrigoyenismo a pluralist route to a modern social state with left
wing autonomous trade unions into a corporatist route.
Uruguay, with a short period of semi authoritarianism in the thirties returned to mass democracy
in the early forties and remained as such until 1973. Again the Colorado party would lead the expansion
of the social state verging on universal coverage already by the late sixties. Yet it is important to note the
mild character of the authoritarian period of the thirties and early forties. Congress, though with the
absence of important parts of both the Colorado and Blanco party remained in full function, trade unions
grew and were not co-opted and repression was moderate, gaining the Terra regime the term
“dictablanda”. The small Socialist Party and the also small Catholic leaning Union Civica were
represented in this congress, and as the transition of the 1940s took place (the good coup of Alfredo
Baldomir) these would be the two parties that pushed and got passed, with the support of the batllista
Colorado’s and part of the Blanco party two bills that would become major building blocks of Uruguay’s
social state and democracy. Indeed, before the Catholic leaning parties in Europe did it, the Union Civica
in Uruguay proposed and got passed noncontributory universal family allowances. Also before de
socialists and social democrats of Europe succeeded, the Socialist Party in Uruguay would propose and
also be able to approve tripartite wage councils, that granted state recognition of public status and the
virtual monopoly of workers’ representation to the left wing trade unions. The answer to the paradox of
Argentina and Uruguay radically distinct political trajectories after the 30s, having had such similar
previous trajectories in the early 20th century, might rest at least partially, on the logics driving their
social state and their products in the late thirties and in their forties. Uruguay never seized to construct its
16
social state based on parliamentary debate and the pressure of left leaning trade unions; Argentina
adopted a verticalist, corporatist arrangement in the expansion of its social state.
Yet either through corporatist arrangements and pushes and pulls of electoral and military
interludes as in the case of Argentina or through stable democracy as in the case of Uruguay, the social
state of these countries continued to expand. The strong migration toward urban centers resulted in the
creation of pressure for the incorporation of subordinated sectors. The State supported this incorporation
by means of expanding public employment and by supporting the domestic industry. The power of the
State over domestic capital, in addition to the early labor organization and unionization of the
subordinated sectors had, allowed the negotiation and implementation of extended programs of social
protection which had a tripartite financing in the case of social security, and State financing in the case of
education and, for some areas, health too. Both the informal and rural population saw only parts of the
benefits of this model. But some factors help us understand why these two countries despite such neglect
are in the 70s the closest we can have to welfare regimes.
First and foremost is the fact that the size of both the rural population and of the informal urban
sector was small. The modern sector of the economy and the State achieved something very close to a
full formal employment between the 50s and early 70s. Indeed, by 1970, close to 80% of the
economically active population was formal. An additional 10% was informal, but this informal sector was
a far cry from subsistence and survival strategies predominant in other countries in Latin America. They
were the petty bourgeoisie of immigrant origin and with family based enterprises (groceries, bars, retail
and other personal services) . These same sectors together with the middle and working classes help also
explain an additional key to the expansion of welfare policies, especially in health and education. Even if
the blueprint of the social state was in these two countries insurance based and stratified, the quality of
public health care and public education was quite high, given the strength of urban popular sectors and
petty bourgeoisie that did not benefit or use insurance based schemes but demanded public noncontributory social services.
The Chilean case is more complicated. It is also a country that had an important Import
Substitution Model development. Its Import Substitution Model was based on mining and on an export
agrarian economy, which was more diversified and had a higher labor force demand. Yet mining allowed
under a modern form of labor capital relation for the emergence of a strong left wing trade union.
Chile also had, like the other cases, an important rural-urban migration, though it occurred later.
yet already by the 1900s important trade unions from Santiago and Valparaiso developed and struggled
through strikes and social mobilization. The first groups to be incorporated to the social protection system
were, like in Uruguay and Argentina, the professionals and public employees. The State elites, using the
17
resources of foreign trade, developed the first programs of social services provision, (education and, to
lesser extent, health) and social security. On the other hand, mining workers that were strongly unionized
and urban workers articulated to left wing political parties in the urban centers, were capable of exerting
enough political pressure and, while not a central banner of the union’s themselves, their pressure
contributed to their incorporation to the systems of social protection and services.
The importance of left wing political parties, of the Christian Democratic Party and of the central
role of congress in Chilean politics cannot be overstated regarding the expansion of the social state. While
formally the right wing parties governed for much of the period between the 1940s and early 60s, the
importance of center catholic leaning parties and left wing populist and programmatic parties, and
especially their behavior, flexible to say the least, regarding their parliamentary and coalitional alliances,
help understand the expansion of the Chilean Social State.
Though it is true that Ibañez is the first architect of the Chilean modern social state, the
continuing strength of the miners, some urban trade unions, the ubiquity of the socialist and communist
parties in the congressional alliances and governments between 1940s and 1970s and the peculiar
ideological and programmatic views of the Chilean Christian Democrats, makes Chile something that
resembles European continental politics more than any other Latin America country. Yet this
resemblance, is in some ways illusory. Parliamentary alliances reflected less ideological positions than
immediate power seeking elites articulated to clientelistic pluralist, yet organized exchange systems
between the parties and employers and workers categories and groups. Thus the Chilean system would
grow and expand through layers and layers of social security schemes. Towards 1970 the Chilean social
security system had more than 100 institutions and programs that covered for different sectors old age,
disability, death, unemployment (only some), sickness and health. Already in the fifties the Chilean
executive had undertaken a study that showed disparities in access, cost and benefits that made the system
unsustainable. Yet the strong pluralist parliamentary and clientelistically based political Chilean logic
made it impossible to reform the system. A mere indication of the pork for barrel politics in Chile can be
seen in the standard duration of electoral coalitions, cabinet members and ministerial postings. According
to Valenzuela (1985) coalitions lasted a few months after elections, cabinets less than year and individual
cabinets posts many times less than a few months. Yet this same coalitional maze and the very pluralist
and institutionalized nature of the game reflected a vibrant democracy by any standard in Latin America.
One in which parties also came together to provide a robust universal educational system, and bargained
and cajoled to expand a stratified but also rather universal and by many standards generous social security
and health care system.
18
These three cases represent, as Huber (1995) pointed out for Chile and Argentina, the
corporative-conservative model that is identified by Esping-Andersen, for the industrialized countries
because of its level of coverage and because of the form assumed by the expenditure. Yet to truly be
compared to the conservative welfare state of continental Europe they should have developed strong
unemployment insurance schemes. While Uruguay did develop some degree of unemployment insurance,
and Chile also did it in a highly fragmented fashion, Argentina lacked one. This weakness brings back the
issue of what kind of social state these were. They were not welfare states, and they were not, to a large
extent because they lacked the idea and political basis to construct some form of social citizenship beyond
equalizing opportunities and providing insurance based schemes for sickness, old age and disability. At
cradle they looked pretty much like a welfare state, close to the grave too, but between cradle and grave,
only economic citizenship, and its link to political citizenship guaranteed some form of stratified social
citizenship.
In any case, and despite these caveats Chile, Argentina and Uruguay represent the closest cases
that resembled in the seventies the welfare regimes of continental Europe in the sixties. Elites that did not
require coercive control of the labor force and that early on had developed a political professional class,
partly autonomous from the landowning elites are critical to understand the early appearance of
education, primary health care and the geologic creation of stratified yet close to universal social security
schemes. Mass politics and especially inclusionary mass politics (democratic in the case of Uruguay,
limited democracy in the case of Chile, populist and corporatist in the case of Argentina) were as nowhere
in the rest of Latin America keys to the expansion in expenditures and coverage of the early southern cone
social states. In the three cases marginal elites and the middle classes formed a reformist alliance meant to
end the oligarchic order (in Uruguay there never was a true oligarchic order, rather a disorder) and
discipline and quiet the increasing vocal trade unions and emerging middle classes. Thus, the early origins
of social policies. After the 30s in Uruguay and Chile these same political professionals, and in Argentina
after the forties a new breed of politicians, be them democratic or not, would carry out the developmental
shift to an import substitution model and would become the shapers of the modern social states. In the
three cases the urban working and middle classes were the majority of the population (less so in Chile)
and played a central role in the process of social state development: co-opted under peronism and in
opposition when the military came to power, and autonomous from state and dominant parties in Chile
and Uruguay.
The form of the social states thus varied but not the depth of such social states. Chile and
Uruguay advanced their social agenda and the social incorporation of the popular sectors combining
clientelism and broader agendas fueled by electoral competition, Argentina advanced it through vertical
19
corporatism and broad policies that were enacted as power changed hands between the military, the
radical party and builder of the social state; the peronist movement.
All in all, what are the relationships that can be posed between democracy and social state in
these stratified universalistic countries.
First and foremost early democratic experiments, even when restricted, are critical in order to
predict early social state origins. Yet, these three countries became in the thirties authoritarian, the effects
of these regimes was radically distinct. In Uruguay it did not alter dramatically the game of politics and of
social state building, in Argentina it brought it to a halt and then allowed for the creation of an unstable
but critical corporatist arrangement that would push for the expansion of its social state, in Chile the
dictatorship of the thirties would actually break the impasse that the conservative elites were placing the
country in, limiting the role of congress and the expansion of vote, and would open the way for the real
origins of the Chilean social state and also open the road for an expanded democratic and still vibrant
parliamentary kind of presidentialism.
Yet the return to democracy, or at least to mass electoral politics would become the building
block in all three countries for the expansion of benefits, services and protections, and in some cases for
the “massification of privilege”. In no country was an organized and articulated labor party with a trade
union constituency the leader of the social state building process. Rarely a major coalitional partner. But
while in Uruguay the space for trade union influence was granted through parliamentary representation
and wage councils at the sector level, in Chile it was through political coalitions and lobbying within
parties, and in Argentina a popular and populist movement either co-opted or repressed the trade unions
to build a verticalist corporatist arrangement. In all three cases democracy and elections increased the
power of these sectors between the 40s and 70s and helped fuel and finance social states.
Finally a relatively clear pattern regarding the effects of social state on democracy. These three
countries presented a stratified almost universal social state, yet while between the 1940s and 1970s one
was fully democratic (Uruguay), another was a stable institutionalized but only increasingly democratic
regime (Chile) and, still another, mobilized, inclusionary but rarely stable or democratic (Argentina).
Well then, is there no effect of social states size, form and coverage on democracy, and especially on
democratic stability? No, at least not in our cases. But there is between the political genetic code of the
social states and the stability and characteristics of their democracies. The argument is thus, once again,
path dependent. Where the origins of the social state was the product of congressional debates,
autonomous trade unions and non ideological catch all parties as in Uruguay, democracy was most stable;
when the genetic code was that of an electorally restrictive political system with an institutionalized
congress, lobbying type trade unions and wide ranging coalitional politics on the left and center,
20
democracy was slow in expanding but stable. Finally when the political genetic code was the result of an
aborted pluralist route recreated and rechanneled in a tense pact between the military, a populist
movement, the organized workers and the elites, where mostly the second and third actors built social
states, while the first and last controlled its limits, the result was a very unstable political system with
many but also short periods of attempted democracy.
As we approach the 1970s these three countries, pioneers in the Import Substitution Model and in
the Latin American social state confronted systemic crises that affected their economy, their societies and
their political regimes. Chile and Uruguay in 1973 and Argentina in 1976 would see their democratic
regimes destroyed. The military regimes that emerged would this time attempt to reshape the
developmental model from the state centered approach to market oriented economies. Regarding their
social states only Chile would radically move in the market oriented direction, Argentina and Uruguay
would financially neglect and at the same time try to put some order into the old social states. Later they
would move more moderately than Chile to the new liberal creed on social policies.
The case of Costa Rica as a particular marriage of democracy and universalism?
Costa Rica stands out amongst Central American countries because of its democratic stability,
relative egalitarianism, GDP per capita and last but not least its social state. Indeed for many analysts
(Huber, 1996; Fleury, 2001) Costa Rica belongs to a class on itself4: the closest case to a universalistic
egalitarian social state or even more an embryonic social democratic welfare state. It is also a country that
has remained export oriented in combination with Import Substitution, and that based its rural production
not on coerced and semi coerced labor, not on militaristic oligarchic regimes, but on a pattern of land
distribution and a form of economic relation that was closer to the farmer model than to the traditional
landed elite model in Latin America (yet for many this last farmer type feature is largely exaggerated,
making the Costa Rican case even more exceptional in terms of its democratic stability and robust social
state).
With the Figueras revolution in the 1940s and the suppression of the Armed Forces, Costa Rica
would initiate a path unique in Latin America. Between 1950 and 1980 Costa Rica would know
uninterrupted democracy, stable growth rates, and increasing social performance. The social protection
system established in Costa Rica in this period was not at first sight so different from those of
universalistic stratification as can be documented by the 19 different special and privileged regimes of
social protection. Yet a closer look would shed a completely different picture on the Costa Rican social
4
Fleury pairs it with Cuba, which is not considered in my analyses.
21
state. First, and maybe most importantly, the Costa Rican state made as axis of its social policy the
expansion of education and primary health care. While the other universalistic countries chose to cover
from top to bottom and tended to add vertically more privileges to the already covered population, Costa
Rica, while also attending this population, rapidly expanded health and education to most of the
population. The fact that trade unions have always been week, and that for a long time two catch all
parties dominated the electoral arena might help understand these broad based policy options. It is also
true that state unionization was always higher than private trade unions. The advantages in protection for
state employees are in some way a direct consequence of this. Costa Ricans like to think of themselves as
a middle class country. Truly, the bureaucracy and independent professionals are “middle class”, truly
also, the old bananero (mostly foreign) and cafetalero elites also exist. But there is indeed a wide group of
people that are neither poor nor middle class and it includes not just urban workers but also small
propertied classes and even workers of the rural areas. They are the backbone of something resembling
though not completely meeting the standards of an egalitarian universalistic social state of basic
protection.
Regarding pensions, once again a closer look shows a remarkable pattern. In Costa Rica between
25% and 30% of the population receives non-contributory social assistance pensions. The 1970s saw the
most important push for social state expansion, and once again coverage of basic income and health care,
as well as education access were the major aims of the governments. Costa Rican society is not a
mobilized society in the sense that Argentina, Chile and Uruguay were. Largely rural and weakly
organized along capital-labor lines, social based solidarity organizations of workers tend to interact with
political parties and political leaders. The absence of strong lobbies articulated to the state (with the
exception maybe of teachers and doctors) has left room for elites that had to gain popular support to move
towards more universalistic goals right from the start. Closely contested elections and wide electoral
participation combined with a relatively autonomous professional state and political elite is the backbone
of the Costa Rican social state.
As Costa Rica’s development model faces the crises of the eighties, a trend towards targeted
liberal type social policies dominated the 1980s governments, yet by 1994, the Figueres administration
would revamp universal social services and attempt to articulate the new small targeted programs with the
fundamental institutions of social protection of the Costa Rican social state. By 1997, health care was
covering close to 90% of the population, and pensions, including noncontributory social assistance
pensions reached almost 82% of the elderly (Mesa lago, 1999). Yet as defined by contributions of the
working population, coverage in 1997 was around 50% (Muller, 2003). The positive gap between
22
coverage of the population and contributions, highlights the importance of non-contributory social
protection in Costa Rica.
A catch all party system in a stable democratic environment and the absence of powerful lobbies
entrenched on the state and the parties, combined with thirty odd years of relative economic growth
contributed to the creation of maybe the rarest jewel in our Latin American social states. Costa Rica
stands today as one of the most socially advanced countries in the continent, and does so with a GDP per
capita that is still quite far from Chile, Argentina and Uruguay. It is also true that Costa Rica lacked any
form of unemployment insurance and that non contributory social pensions only amounted to half the
basic insurance pension (Mesa Lago, 1999). Still for a country that in the forties was overwhelmingly
agrarian and lacked almost any form of social protection and services (with the exception of education)
what has been achieved is no small feat. It is also unthinkable to speak of the Costa Rican democracy
without considering the role that the social state has played. Through education first and through health
care and social security later the Costa Rican state rests on the solid grounds of basic social citizenship.
The allegiance of Costa Ricans to democracy is not naive, it is based on the proven conviction that elites,
elections and society can find middle grounds in the construction of a mixed economy with social
protection as one of its main building blocks.
Maybe in no other case, is as clear as it is in Costa Rica, the relevance of democracy as an
independent variable pushing social state expansion. While this country is a relative latecomer to modern
social policy, it is between the 1940s and the 1970s, a period of unprecedented democratic stability that it
develops and institutionalizes its social state. Costa Rica, a coffee based and fruit based, export economy
in the 1930s, had few structural factors that would predict a robust social state. Without the Figueres
revolution, the suppression of the armed forces, and the creation of a stable two party system that had to
compete for the popular vote, it is hard to even imagine how this small Central American country, could
boast today, of what some define, as the only egalitarian Latin American Social State.
ii. Dual Regimes based on elites statecraft and cooptation and repression of the popular sectors
(Brazil, Mexico5)
These regimes had, until 1970, nearly universal development of primary education and an
important, though stratified, degree of health coverage. Concerning social security, the coverage
5
The cases of Colombia, Venezuela y Panama could also be classified under the dual types yet I lack systematic
historical and statistical evidence for a proper analysis. The case of Peru is analyzed below as a peculiar form of
dualism.
23
accentuated the stratified aspects that characterized mature systems without presenting the universalized
coverage of the stratified universalism regimes.
In these countries the problem of territorial heterogeneity is clearer. In some provinces and
regions, there is an important development of the formal market, the state and social protection while, in
other provinces, there is a virtual exclusionary system with very low incorporation of the vast majority of
the population (either through the market or through the State).
These countries have been characterized by having an Import Substitution Model development
supported based on rents coming from primary economies of high labor force intensity and with an
important regional heterogeneity in what refers to levels of economic and social development.
Both countries developed and matured their social states under what has traditionally been
considered as a populist model of development and political administration. Cardenismo in Mexico and
Varguismo in Peru represent this particular form of political and social incorporation. Mass movements
and mass politics was indeed a characteristic of these models. In the case of Mexico the roots of mass
mobilizations and politics lie early in the Mexican Revolution and have a peasant based as well as urban
based constituency, while in Brazil the impulse from below comes essentially from the modern working
and middle classes of the great urban centers, and even then would be rapidly controlled and co-opted
from above.
TABLE 3
Selected variables on social policy effort for dual welfare countries, circa 1970
% covered %
by Social by
covered Social
Social expenditur under one year primary
Security in Security
EAP
% of children % enrolled in % enrolled in
in e
as
secondary
% vaccinated
education
of education
age group
age group
total
GDP,
against
population
1980
tuberculosis
Brazil
27.0
47.0
9.3
66
82
26
Mexico
28.1
53.4
8.0
80
104
22
of
Sources: Mesa-Lago, 1979, 1991; World Bank-World Development Report-, 1994, Miranda, 1994,
Cominetti, 1994; .
24
From the political point of view, the control and incorporation of popular sectors have rested, in
areas of lower economic and social development, on a combination of coercion and local elite’s
patrimonialistic forms while, in more developed areas, they have assumed the shape of vertical
corporativism, articulating trade unions, parties and state. In Brazil during the Varguismo era this
assumed two distinct political forms: the so called social democratic party (PSDB) that at the rural areas
operated as a form of state-local elites-party control closely resembling the old patrimonial politics, and
the Worker’s party of Brazil, oriented towards the control of labor, through trade union, party and state
articulation (Draibe, 1985). In the case of Mexico the revolutionary party with and after Cardenismo
would include within the same party the “Confederación Nacional Campesina” and the “Confederación
de Trabajadores de Mexico”. Yet while in Brazil the PSB was mostly a new name for old forms of control
the CNC in Mexico was a new form of control a postrevolutionary model, which had its autonomy and
power especially during the early Cardenismo. After the 1940s many of the institutions and leaders of the
CNC would be relegated to the more urgent concerns and vocal capacity of the CTM (Brachet, 2004).
Yet in both cases, the frailty, short duration and lack of intensity of electoral competitive
democratic regimes allowed for the development of this dual form of incorporation, where we could
witness increasing differences between the power of rural and urban sectors, backward and advanced
areas, and formal and informal labor. Unlike stratified universalism countries, the social state moderates
(yet, having a stratified nature) the social segmentation only in those sectors that are incorporated into
modern formats of protection. However, the protection systems accentuate social differences between
urban advanced sectors and the non-protected rural traditional and urban informal sectors. Furthermore,
one of the main features and future challenges of these systems is a direct legacy from peculiar form of
statecraft (Malloy, 1998) that these countries pursued.
The state, in these regimes, became the building block of the restricted populist coalition that was
in the basis of the labor incorporation period. The state was not merely, as it was in the universalistic
countries a source of resources that contending elites used as a basis of competition for popular support
and market regulation. The state was the realm were a previously patrimonial, traditional and status based
political and social order found its ultimate expression and refuge from the more meritocratic logic of
modern markets, thus creating a state apparatus less bent towards redistributing for popular support and
more oriented towards redistributing for elite accommodation and popular control. The wages and
pensions of state officials are indeed extremely high in these regimes, far higher than those of Chile,
Argentina and Uruguay, when compared to the private sector. Still today, in Brazil, approximately
1.000.000 pension beneficiaries from the state take up a similar amount of money as 14.000.000 pension
beneficiaries from the private sector. Still today, too, as Heredia (2002) points out the Mexican State is
25
unique not so much because of its size but because of how much the state employees wages represent of
total state expenditure, almost 45% of it, the highest by far in Latin America, while in terms of percentage
of state employees over total working population Mexico ranks about average in Latin America.
TABLE 4
Selected social development indicators for dual welfare countries, circa 1970
%
Households
below
of %
of Illiteracy rate
Households
the below
Infant
Life
Mortality
Expectancy at
the
birth
Poverty line
indigence line
Brazil
49
25
33.6
95
59.8
Mexico
34
12
25.8
60
62.7
Sources: ECLAC -Panorama Social de América Latina- 1995; ECLAC Anuario Estadístico de América
Latina- 1980.
The data on social indicators in Mexico and Brazil, present, despite their commonalities, important
differences. Mexico, clearly, out performs Brazil in all social aspects considered here.
The most
important reason behind such differences cannot be placed on social policy spending. As a matter of fact
Brazil spent more by the late seventies and early eighties than Mexico did. Three differences do help us
understand dissimilar social outcomes. First the overall pattern of inequality, much more entrenched in
Brazil than in Mexico, especially in the rural areas, where the latter had an agrarian reform based on a
revolutionary movement and on armed peasant struggle. Secondly the fact that social spending in Mexico
concentrated more on health care and education than it did in Brazil, where the emphasis was on the
pension system. Finally the fact that in Mexico, even though as a weaker partner, the rural population and
it corporatist organizations were a part of the populist pact, and indeed a major building block of the PRI
structures. As Brachet (2004) points out, Mexico starting in the 20s and thereafter with Cardenismo
incorporated to a larger extent than other dualist social states the peasantry and poor rural population as
not merely “issues” but also as subjects:
“Since the 1920s, once the country had returned to peace and a new regime had been established under
solid constitutional grounds, the most innovative (as a matter of fact revolutionary) initiative regarding
26
the social issue was the transformation of the poor and indigenous peasant into a visible subject with
rights, first to land and education and afterwards with Cardenas to health.”6
In contrast, Brazil was as Draibe (2004) puts it, maybe the quintessential “industrially biased”
developmental social state. By such she means a social state that centered on regulating industrial
relations between labor and capital and providing protection only for labor that was engaged in such
modern labor-capital relations (and of course public employees). This industrial bias anchored in laborcapital relations also helps explain why Brazil had such emphasis on social security leaving education and
health-care neglected for a long time. In her words:
“The same “industry bias” seems to have magnified the place of social security policy in the Brazilian
Welfare State, leaving small or marginal room for all other social policies. Although the reformist
wave of 1930 had affected practically all social areas, through new national codes and a copious
legislation, evidences show that social policies such as education and still more the health policy, have
not occupied a central place in the governmental social agenda, nor in the successive development
plans from 1930 until 1964”
Despite this very important differences, Mexico and Brazil, remain more alike than different, and the
major reason once again has to do with the capacity of the modern economy and the state to incorporate
people into “economic citizenry”, or to formal protected labor-capital relations. Indeed by 1970 Brazil
had 52% of it economically active population in the formal sector, while Mexico had 57% of it in the
formal sector (Isuani, 1984). Thus, roughly half of the country was “in” while the other half was “out”.
The main difference here is that while the half that was out in Brazil was excluded not only of the formal
sector, but also of social security and health (and almost 20% out of primary education), in Mexico,
primary health care and basic education (although of a very low quality compared to those in the formal
sectors) had in one way or another reached most of the population. Thus mass politics and institutions
help us explain a major variation within similars: they were essentially dual regimes, yet in Mexico a
urban-peasant revolution and a party that explicitly and really integrated the peasant’s organizations into
the populist party, stands in stark contrast with the almost cosmetic creation of a national party in Brazil
(the PSB) that was essentially a new name for the traditional “coronelismo” practices of the landed elites.
The effects of such differences were the “industrialist bias” of Varguismo and the “limited space” but
6
Own translation; A partir de los 1920s,una vez pacificado el país y establecido el nuevo regimen
sobre una sólida base constitucional, lo más innovador (y para la época realmente revolucionario) en
materia social fue la transformación del campesino pobre y/o indígena en un sujeto visible y con
derechos: primero a la tierra y a la educación, y posteriormente (bajo Cárdenas) a la salud”
27
space nevertheless for rural oriented social policies, especially in health and education in the Mexican
case.
In any case the interaction of democracy and social protection in these countries and to a certain
extent in Argentina was weak, based more on narrow lobbying, cooptation and result based legitimacy
rather than procedure based legitimacy.
The worst effect on democratic culture and values that these regimes had were not only the limits
they imposed on social incorporation, but the effect they had especially in the urban highly mobilized
population in the creation of an extended belief system based on the idea of corporatist rights rather than
citizenship rights, and a tendency to legitimize governments and even regimes based more on their results
than on their procedural respect of formal democratic rules.
The case of Peru: from dual to exclusionary?
The case of Peru deserves special attention for it is a country that had developed a format of
political incorporation and indeed a party that represented such format (APRA) that could have well led to
a dual model of incorporation and eventually to a more integrated type of regime. Yet somewhere along
the path, Peru started to show signs of becoming increasingly exclusionary. The ethnic and regional
cleavages, the role of the Lima and rural elites and the failure, first of APRA and then of the military
experiment to gather and structure a stable alliance between industrial labor and middle classes with elites
tolerance helps us understand the outcome as we enter the early 80s.
Indeed between the late 1930s and the late 1940s Peru walked the path of other dualist regimes.
While only asystematic measures (such as the health campaign vaccination of the Rockefeller foundation
and the education expansion in rural areas doomed to failure because of linguistic barriers) were taken
for the rural population, at the urban centers and in the mining and intensive labor agricultural enclaves
organizational capacities developed and the APRA emerged as a powerful articulating force. The Prado
and Bustamante governments of the late thirties to late forties were receptive to the new party and to the
new social forces expanding social security benefits to the middle and working classes7 and developing
7
Between 1936 and 1948 benefits on Old Age Disability and Survival were passed for Public employees, port
workers, sickness and maternity benefits for industrial workers and private employees. Even taking into
consideration that only 30% of the population was urban and only a part could benefit from these new systems, the
basis were being laid for a dualist type of regime. Yet between 1950 and 1961 neither one new benefit nor one new
group was incorporated into the system with the exception of additional benefits for the armed forces in 1950 (Mesa
Lago, 1985).
28
wide ranging price controls. In one decade state spending grew almost 5 times and export taxes more than
tripled to meet these fiscal demands (Sagasti, Prada y Bazan, 2004).
Yet this incipient road to urban stratified protection was cut short by the coup of general Odría,
who backed by the economic elites both national and foreign who feared the APRA, moved away from an
interventionist state to a liberal one. Thus one of the major peculiarities of Peru´s trajectory is that while
most of Latin America was embracing in the 1950s import substitution, Peru was in an open strategy
based on laissez faire principles. These had two devastating effects on the chances of Peru´s social state.
The single most important effect is that Peru did not have the ISI push to expand in a dual economy the
“modern” part of it especially in terms of employment creation. This does not mean that modern
economies did not develop, but the dual character was intensified. As Sagasti, Prada and Bazán (2004)
point out:
“While most Latin American countries succeeded in strengthening their industrial base and internal
markets as a response to the economic crisis of the 1930s, Peru enjoyed a limited success due to the
structure of the economy, which was composed mainly of mining and agricultural foreign enclaves; and
the political domination of a landowning elite. Unlike many Latin American countries, where the 1930s
and onward was characterized by the emergence of a developmental state, this did not fully materialize in
Peru until two decades later...”
The social state, thus also was restrained. Most social policies from that time are restricted, though these
by no means implies they were irrelevant, to urban infrastructure, housing and education. Income
maintenance schemes and Bismarckian type protection had little if any space both in the Odría regime and
the second Prado administration. By 1961 only 8.6 % of the total population were covered by any form of
social security. In the 1960s a new government now headed by Bealunde would launch an aggressive shift
in development joining, thus, Peru, its Latin American peers in Import Substitution Industrialization. But
this was not accompanied by a drastic increase in social security coverage. Indeed only by the 1980s
would social security in Peru reach 17% of the total population and slightly more than 30% of the
economically active population.
Thus the case of Peru highlights the interdependence of the lack of stable political regimes and
cumulative advances in social protection. While Mexico and Brazil faced the problem of corporate
incorporation, Peru faced the challenge of truncated corporatist incorporation. The weakness of
democracy and indeed of inclusionary mass politics in Peru, is to a large extent the mirror image of its
weakness in developing stable, even if limited, social protection systems.
29
iii. Exclusionary Regimes based on predatory elites. (Guatemala, Honduras, El Salvador,
Nicaragua, Bolivia)
The Central American countries
and, probably Bolivia and Ecuador, have elitist systems in what
concerns social security and health and a dual system referring to education. In these countries, poverty,
measured according to income, exceeded in all cases 50% of the population . Toward 1970, less than
20% of the population was covered under basic programs of protection and social security, except
Guatemala that had 27% of its population covered. Concerning health, the reality was similar. However,
in what refers to education, primary education was expanded, with strong quality inequalities, reaching
more than half of the population and, in some cases, reaching almost universal coverage.
TABLE 5
Selected variables on social policy effort for exclusionary welfare countries, circa 1970
%
% covered Social
%
covered
by
children
by
Security in as % GDP, under
Social
total
Social expenditure
1980
of % enrolled in % enrolled in
primary
secondary
one education
of education
age group
age group
year
Security population
vaccinated
in EAP
against
of
tuberculosis
Bolivia
25.4
9.0
5.9
70
76
24
Ecuador
14.8
9.4
9.3
61
97
22
Nicaragua
14.8
9.1
2.3/s/d(a)
90
80
18
El Salvador
11.6
6.2
1.3/1.7(a)
s/d
85
22
Guatemala
27.0
14.2
1.6/3.7(a)
--(b)
57
8
Honduras
4.2
7.3
0.9/12.2(a)
75
87
14
(a) Social security expenditure/Health care expenditure; (b) vaccine is not given.
Sources: Mesa-Lago, 1991; World Bank-World Development Report-, 1994, Miranda, 1994, Cominetti,
1994.
30
These cases are quite close to what Peter Evans (1992) denominates Predator States. In such
countries, elites appropriate themselves of the State apparatus and, supported on the export of primary
goods in enclave economies, use the fiscal capacity of the State for the extraction of rents, without
providing the counterpart collective goods neither in infrastructure nor in regulation or social services.
Protection and social security systems of this kind consist, in great extent, in elitist policies that add
additional privileges to a group population that already enjoys a privileged situation. Professionals, a
very reduced number of formal workers and public officers are those that, typically, benefit under
these models. The vast majority of the population (represented by the informal sector workers,
agriculture and secondary labor force) is excluded.
It should be pointed out that these countries are characterized by a high heterogeneity in what
refers to social structure. This heterogeneity is reflected in the differences of distribution of the national
wealth between urban and rural sectors, between different rural areas and between urban centers and their
periphery . In other words, the degree in which the market itself operated, generated the creation of
“clusters” of negative and positive social configurations. From the State side, the degree of intervention as
“risk pooling” has been minimum.
Likewise, the State bureaucracies reflect the same situation. They also have a marked differential,
in what concerns efficiency and productivity, between privileged areas of the State at the central level
and sub-national dependencies of less importance. In some cases, we face the lack of state bureaucracies
at regional levels.
Finally, at the political and social level, the organized participation of subordinate sectors in the
decision making, has been historically low due to the predominant repressive character of the political
regimes and the low organizational capital of these sectors. The control and co-optation of the subaltern
social sectors have been reached at best by forms of clientelism and patrimonialism and in its worst
versions by sui-generis forms of repressive and authoritarian semi military actions and structures.
Consistent with this realities, in these type of countries, the social indicators systematically have the worst
figures as well as the highest differentials between regions of diverse degrees of development.
TABLE 6
Selected social development indicators for exclusionary welfare countries, circa 1970
%
of %
Households
below
of Illiteracy rate
Households
the below
Infant Mortality
Life
Expectancy
the
birth
31
at
Poverty line
indigence line
Bolivia
50(a)
22(a)(b)
37.3
157
46.7
Ecuador
s/d
s/d
25.8
100
57.1
Nicaragua
s/d
s/d
42.1
109
52.9
El Salvador
s/d
s/d
42.9
92
59.1
Guatemala
65(c)
33(c)
53.8
104
54.6
Honduras
65
45
40.5
110
54.1
(a) 1981; (b) Only urban areas; (c) 1980.
Sources: ECLAC -Panorama Social de América Latina- 1995; ECLAC Anuario Estadístico de América
Latina- 1980.
The rural population and most of the urban population is essentially unprotected in these systems.
Only education making some inroads into the predatory logic. The history of at least some of these
countries has changed dramatically, though not yet their social realities. Revolutionary rural based and
urban led guerrillas in Nicaragua and El Salvador brought down or drastically changed the nature of the
regimes. While still patrimonial, clientelistic and indeed many times repressive, the governments of these
two countries do have some accountability and these changes might in the future have a positive impact
on social policies. If dualist countries advanced in a peculiar form of conservative modernization regimes
that while not democratic were at least inclusionary of urban working and middle classes, the democratic
struggle, at least in some of the exclusionary countries took the form of revolutionary armed movements,
for no other avenue of change was open to them.
CRISIS OF THE OLD MODELS AND THE NEOLIBERAL TURN
a. The economic and social critique of Import Substitution Industrialization
32
In January 1990 the Institute for International Economics published Williamson’s “The Progress of
Policy Reform in Latin America”. In it, the author put forward the idea of the “Washington Consensus”, a
set of policy prescriptions advocated by multilateral institutions and US authorities regarding the
development process in Latin America. His views as well as additional contributions in the volume
breathe a sense of accomplishment and celebration, though not without notes of caution. Williamson sees
a region that ten years earlier had oversized and inefficient states and an inward oriented development
strategy radically changed, and changed for the better. While he recognizes the 80s as the lost decade for
Latin America in terms of growth, he argued that major transformations in the role of state and markets
promised a significantly better decade.
Williamson’s contribution does not ask why reform took place nor why reform followed certain
recipes. He is neither a sociologist nor a political scientist. His concern is simpler: Did reforms take
place? Yet, behind such a narrow concern rests a formidable faith, namely, that those were the right
reforms. The “Washington consensus” is not merely an agreement, it is the right one. Part of this faith is
based on an additional belief: the absolute conviction regarding the ills of the previous development
model, (i.e., import substitution industrialization). In effect, import substitution is widely regarded today
as Latin America’s wrong turn in development. From an economic point of view, critics insist on the
detrimental effect of trade barriers on Latin America’s natural comparative advantages. As for the role of
the state, it is argued that internal regulations and oversized and inefficient bureaucracies inhibited the
role of the market in allocating resources efficiently, and contributed to the region’s chronic fiscal deficits
and inflation prone economies.
In addition to this critique, more sophisticated versions based on a political economy approach
argued that the model in and of itself favored the creation of a rentier behavior on all parts involved in
economic activities, thus inhibiting capitalist development. The argument is relatively straightforward in
its most stylized fashion. The state subsidized the import substitution elites from rents it extracted from
international trade taxes. Trade unions asked for increasing benefits and wages that were promptly
granted for political reasons as well as a strategy to minimize labor conflicts. In turn, these increases were
passed along to the consumer in the form of higher prices, which was possible given the fact that the state
protected their market space and share through the granting of tariffs, legal monopolies, tax exemptions
and credits and subsidies. Capitalists, labor and labor unions had no incentive to favor strategies that
increased productivity and shared the costs and benefits of innovations since they had an easier way to
maintain profits and gain wage increases. This state of affairs was defined by Evans (1988) as the creation
of an anti-Schumpeterian triangle were protected workers, protected entrepreneurs and the state elites and
33
employees engaged in rentist games fueled by a political logic and not by the driving force of competition
and innovation. This was, then, a zero sum game, which in the long run would turn into a lose-lose
situation. The country as a whole would pay the price of decreased productivity in the non-tradable sector
(thus never developing an internationally competitive industry) and increasing added costs to the tradable
sector, deterring investment further and placing incentives on rentist strategies.
The complement of the economic critique is the widely held notion that Latin America’s
abominable pattern of inequality and dismal record of social development are also due, at least in part, to
this import substitution model. The argument is somewhat more complex and more shaky than the
economic critique of Import Substitution Model. The social critique can also be framed in what Evans
termed the “anti-Schumpeterian triangle” of State, Urban Labor and Industrial Elites. The Import
Substitution Model redistributes income within this triangle but in the process deprived three groups or
social sectors of protection and benefits.
In the first place, it clearly hurts landed elites that have to part with some of their profits in order to
finance subsidies and also have to accept an exchange rate that diminishes the competitiveness of their
products. This group, in contrast to the industrial elites protected from outside competition by the Import
Substitution Model is a price taker in the market and thus cannot pass its losses to the consumer or buyer.
Secondly, rural labor and small agrarian propertied classes lose. The burden of countries’
development rests in this model within the tradable sector that in most of the region was based on a
couple of agrarian staples. Consider first the case of medium and small proprietors who sell their product
to larger export conglomerates. They are the weaker link and the ones who will carry most of the weight
of this burden. Consider now rural labor. Again, elites can only pass onto them their costs in the Import
Substitution Model. This happens through the impact that rentist behavior and low investment have on
agrarian labor, wages and employment.
Finally, consider a third group: informal urban labor. These workers are not protected in the Import
Substitution Model They are indeed the surplus labor the model could not incorporate with their rate of
industrial development and state employment. While they do not benefit from the protection and subsidies
that both formal labor and entrepreneurs enjoy, they do suffer the increases in prices that this model
generates by passing increasing wage costs into commodities prices. The end result is that these societies
end up dual, with a small protected domestic modern sector and a large unprotected rural and informal
urban popular sector. Based on this diagnosis, MLA and market oriented analysts claimed that the system
of social policies and labor protection in Latin America had a large responsibility not only in curtailing
the region’s economic growth, but also in undermining its social development. In order to both improve
34
economic performance and better social development, a new social policy paradigm emerged for Latin
America.
b. The regional social policy agenda: structural forces, political dynamics and crossroads for
the Latin America Social State.
The “Washington Consensus” avoided the debate on social issues and social sector reform as issues
that were beyond the scope of structural economic reform (Williamson, 1990). After the Washington
consensus we have witnessed the attempt by MLAs, the US governments and part of Latin American
domestic elites, to develop a new Washington consensus regarding social policies. If the original
consensus concerned itself with economic reform of the liberal kind, social sector reform chooses similar
paths and contents, and is basically supported by the same doctrines and theories that underpinned the
economic market revolution that Latin America has undergone in the last twenty years.
These turn of events and renewed interest in social development and social policies responds to a
number of reasons. Social development has stagnated or critically slowed down in most of Latin America
in the last decade and a half. Consistently, social protests and turmoil have increased significantly. Also,
there is a widespread perception of the inadequacy of the traditional model of social policy in Latin
America, and thus the need to reform it. Thirdly, social policy has achieved a new status, one that fits
better the dominant economic ideology and the economists’ hegemony over the social sciences and the
issues of development in general: certain social policies have become forms of investment, rather than
simply expenditure. We now make social investment in human capital instead of simply providing
services to those in need or redistributing income. In any case, this new found interest of the international
community in social development and people’s welfare should be welcome after too many years of a
narrow economic focus regarding development strategies.
With this renewed interest in the social question came the critique of the old model of social
protection. Traditional social policies (those associated with the import substitution model) have been
criticized as inefficient, expensive, unfair and rigid. This critique pointed out that the supposedly
universal system of social services never reached those truly in need. It also highlighted the excessive
administrative costs of state run social provision, its centralized nature with its effect on incorrect
resource allocation, and its clear-cut stratified nature in the coverage, range and quality of social services
and transfers.
Structural forces at play further exposed the inadequacy and costs of previous traditional systems
and made them both more acute and visible. In the first place, economic openness placed new strains on
35
entrepreneurs’ need for competitiveness and increasing costs on economies that run fiscal deficits, high
and unstable rates of inflation, and overall macroeconomic instability. These factors in turn pressed
towards lowering the fiscal burden of exporters and at the same time keeping the strictest possible fiscal
discipline in a time of few state resources. This implied among other things controlling social
expenditure, and prioritizing fiscal discipline and lower taxes over social needs.
Secondly, and especially in the most “modern social structures”, the mature social security systems
had to cope with the aging of the population which meant that less people were working to sustain the
lives of more older people that were not. At the same time health costs skyrocketed as older people
required longer and more expensive treatments for cancer and cardiovascular diseases. Also, the
educational system had almost complete primary level coverage, pushing for the expansion of secondary
and later tertiary education at the same time that it increasingly presented wide disparity in quality even
within the public systems.
Finally, and may be most importantly, both unemployment, and informal employment grew in
almost all countries of the region.
TABLE 7
Open Unemployment in Latin American Countries, 1990-1999.
Argentinaa
Bolivia
Brasil
Chile
Colombia
Costa Rica
Ecuador
El Salvador
Honduras
Mexico
Nicaragua .
Panama
Paraguay
Uruguaya
Venezuela
1990
5.9
9.4
4.5
8.7
9.3
5.3
6.1
9.9
6.9
3.3
1994
13
3.2
7.4
6.8
8
4.2
7.1
6.8
4.1
4.5
14.1
15.7
4.4
9.7
8.9
18.6
6.3
8.9
10.2
1997
14.3
3.7
8
6
11.8
5.8
9.2
7.3
5.2
5.1
13.1
15.4
8.4
11.4
10.6
1999
14.7
7.1
11.4
10.1
19.2
6.1
14.2
6.9
5.3
3.2
13.8
13.1
10.1
11.2
14.5
Source: CEPAL, Panorama Social de América Latina, 2001-2002. a. Only urban areas
A social security and health care system based on wage contributions proved to be, again, as it have
proved for most countries before, extremely limited as a means of providing social protection to wide
36
sectors of the population, especially the poor. In short for the new reformers the Latin American Social
State of the past created rather than fought poverty and inequality and did so while at the same time
spending too much and doing it inefficiently.
As we have shown this diagnosis lumps together very different realities.
Indeed, a central error of the overall critique of the social aspects of the Import Substitution Model
is that it does not take into consideration the important variations it presented throughout the region. In
effect, not all countries developed the social side of the Import Substitution Model equally. In other
words, the size of the anti-Schumpeterian triangle varied widely within Latin America. While the Import
Substitution Model critique might very well be valid for cases in which the triangle reached less than a
third of the population, it is less so in cases in which it reached more than half of it, and clearly without
merit as we move to the few cases in which most of the population benefited from transfers from the
landed elites and the export sector to the urban protected sectors. While dual social realities represent the
end result of a good part of Latin America, Central America ended up not dual but exclusionary, precisely
because the ISI model was weakly advanced, not because of its strength. Finally the southern cone
countries did not end up dual but rather universalistic, though they did end up highly stratified in their
system of social benefits and protection. Both cases of dualism and universalistic stratification played a
central role not simply in protecting their population but also in building democratic and semi democratic
regimes. The political process of incorporation of the popular sectors and the middle classes is to a large
extent the social process of incorporation of these sectors, into education first and social security and
health care later.
Yet a part of the social critique of ISI makes perfect sense. Overall our social states were inefficient
in giving basic protection to its citizens. Inefficient, but some of them not necessarily ineffective. Highly
stratified social insurance schemes based on contributions from wages, health care for the middle classes
also based on stratified differentiated systems of compulsory and voluntary insurance, and the absence of
a strong tax base to support non-contributory health care and education made for inefficient, deficit prone
and quality stratified social protection systems. Yet, while this is true, the road open for reform is not self
evident. Latin America opted both in its new economic model as well as in its new social protection
system overwhelmingly for neo-liberal solutions or what Sonia Fleury (1994) has termed the
liberal/productive model. Why this was the preferred choice in most countries responds not mainly to
technical criteria, nor just to the declared intentions of equitable and sustainable growth of some of its
proponents, as honest as they might have been. The reasons are sadly more endemic to Latin America
37
policy making and place a big interrogation sign on the quality and sustainability of our regional
democratic processes: fiscal stress, national and international elites narrowly understood self interest,
avoidance of immediate domestic political costs by parties and party leaders, and weakness of a broad
based political movement from below.
Yet the road to liberal reform has also opened up technical and eventually political possibilities for
a new type of social state: one geared at universal coverage of basic rent, education and health. In a way
this is not surprising. Esping-Anderson had already documented that liberal and social democratic states
where historically related, while corporative conservative, Bismarckian type welfare regimes tended to
stand alone.
AGENDA, REALITY AND POLITICS OF SOCIAL SECTOR REFORM AFTER THE 1980s
Both the diagnosis and the international and internal forces pushed for social policy reform.
The content of such a reform should, in the eyes of those who criticized the old model, be able to
avoid the problems and inefficiencies of traditional social policies. In order to do so, certain basic
features of what the new social policy system should look like slowly took shape in the minds and
documents of market oriented social reformers: targeted social policies, decentralized administration
and decision making, private providers both for profit and non for profit, and a shift from supply based
social policies to demand based social policies (or in other words from financing supply to financing
demand).
The menu for reforms is today a matter of debate, but it truly represented a new Washington
consensus in the 1980s and especially in the early to late 1990s. Such agenda was different in its
liberal stress and accomplishments depending on countries realities and politics as well as policy
sector specific configurations.
a. The late 1970s and 1980s: A pure pioneer, parametric reforms and social state erosion.
i. Fiscal bankruptcy, democratization, social emergency funds and the fall in social spending.
The 1980s has been labeled the lost decade for Latin America. It was preceded in the 1970s by the
increasing awareness of the ISI model exhaustion. In most Latin American countries the late seventies
and a good part of the eighties were also times of dictatorship . A simple map shows the rather gloomy
political realities of Latin American Countries between 1975 and 1980.
38
FIGURE 1
Combined with the crises of the ISI model came the debt crises of the eighties and authoritarian
regimes. The major impact of this configuration of forces was a radical neglect of social policy. The neoliberal agenda of the eighties is not particularly centered on social policies. While there was, from the
multilateral agencies a recognition of the need for what were called “emergency funds” meant to deal
with the harshest social costs of adjustment, not much was done regarding the old structures of the
developmental social state. Emergency funds are a direct result of the social costs of adjustment and also
39
a response to the democratization processes of the eighties in a context of fiscal bankruptcy. For many
these emergency funds were meant to help democratic governments advance adjustment and structural
reform at the same time that they would be seeking support in the poorest sectors of society.
Also these emergency funds were already based on a basic tenet of what would later be a more
articulated neo-liberal view on social policies: the role of the state and of social policies should be kept to
a minimum, ideally it should also be temporary and only targeted to those sectors of the population that
were poor. Given the levels of poverty in dual and especially exclusionary countries, the expression
inevitably became “to the poorest of the poor”.
There is a telling pattern regarding the emergence of these social emergency funds: the least
developed the social state of the past the more and more fragmented effort in terms of emergency funds.
Indeed while Guatemala had 6 emergency funds in the eighties, Ecuador had thirteen, Honduras had five
and Peru had seven, Costa Rica , Uruguay, Chile and Argentina had one or two at the most (Filgueira,
1997).
The case of Mexico with Pronasol is also an interesting example of the problems this approach
presented. A study by Heredia (1990) showed that Pronasol invested most heavily not following poverty
but following areas where the PRI was facing electoral hardship. The highly discretionary character of
these funds, while limited the old complex corporative access to them, allowed for an even more pure
form of clientelism and co-optation.
As adjustment and emergency funds developed, the old social state was neglected. Indeed, the first
impact of the crises and the neo-liberal agenda on Latin America social states was not structural reform,
but reform by default. That is reform by deterioration. If we consider the average per capita expenditure
and the fiscal priority of all Latin American Social States at the start of the eighties and compare it with
the mean of the period from 1982 to 1989 we will clearly see this pattern.
Fiscal priority of social spending (social
spending/total spending) in Latin America
70
60
50
40
30
20
10
Fiscal Priority 1980-81
40
Bo
liv
ia
Ur
ug
ua
A
rg y
en
Co tin
st a
a
Ri
ca
Ch
ile
Br
a
V
en s il
ez
ue
Co la
lo
m
bi
a
0
Fiscal priority 1982-89
Source: ECLAC, 2000, 2001, 2002
In all countries to a larger or lesser extent the proportion of social spending as a percentage of total
spending fell. If we divide countries by their previous levels of social spending into high, moderate and
low spenders8, we can see that in per capita terms the realities of the eighties were equally distressing.
Even those countries with a more robust social state saw their spending fall, in some cases dramatically9.
Per cápita expenditures in high, moderate and low
social expenditures countries
350
300
250
200
150
100
50
0
High social spending
countries
Moderate social
spending countries
Per capita spending 80-81
Low social spending
countries
Per capita spending 82-89
Source: ECLAC, 2000, 2001, 2002
When an analyses is performed on the impact that democracy and democratic transitions had during
the eighties on social spending seen both as fiscal priority and as total percapita spending the answer is
straightforward: “democratization and democracy did not lead during the eighties to a protection of social
spending” (Avelino, 1997).
ii. Parametric and homogenization10 reforms in the universalistic stratified models and the
pioneer market oriented case of Chile.
8
High social spending countries include in the 1980s Argentina, Uruguay, Chile, Costa Rica, and Brazil. Moderate
are Colombia, Venezuela and Mexico and low are Peru, Paraguay and Bolivia.
9
Argentina saw its per capita social spending go down from 569 dollar to 470 and Mexico from 224 to 163.
10
Parametric reforms refer to changes in the parameters of social security benefits (age of retirement, replacement
criteria for defining benefits, time of work for eligibility, etc). Homogenization reforms refer to parametric changes
that sought to unify parameters across a wide range of differentiated social security schemes in each country.
41
By 1970 Uruguay had more than a 100 “cajas” or pension systems, and Chile was close to sixty,
Argentina presented a similarly fragmented model. The authoritarian regimes of the 70s attempted, based
on their new gained autonomy from social and political groups, to unify and simplify these systems. As a
matter of fact, technocrats inspired in the ILO principles, from the social security administrations thought
that it was possible, now that corporations were at least partially out of the picture or had less leverage
with the executive, to start grounding the principles of uniform treatment and solidarity long forgotten in
the southern cone systems. Both Uruguay and Chile were able to drastically modify the system within its
own logic. Retirement ages and replacement criteria were unified and a large part of the old “cajas” were
brought under unified authority and administration. Still many segmentations remained, and in the case of
Argentina little advances were made. At the same time these movements towards homogenization were
accompanied by less generous parameters across the board, regarding eligibility criteria, replacement
rates and range of protection.
Yet the increasing influence of ILO technocrats within the state, gave way in Chile to another group
of technocrats seated in the areas of finance and economy: the Chicago boys. They had a more radical
idea of how to deal at the same time with heterogeneity of parameters and generosity of parameters, and
thus with deficit in the long run. The capacity of these technopol elite to transform the Chilean state
knows little precedents in Latin America. Under their advise, in 1981, Chile would become the first
experiment in a new breed of social security systems: mandatory individual capitalization funds
administered by for profit private organizations. The law was passed without consultation and the
administration of these funds was given to alliances of the big Chilean financial conglomerates and
international capital.
For many such a drastic reform of the basis of the social state was only possible because of the
authoritarian regime. Its effects implied the destruction of a dormant but nevertheless existent old network
of influences and relations between trade unions, party leaders, and state officials. As we shall see the
prediction that such reforms would not be possible under a democratic regime proved wrong. Even
though Chile remains as the purest example of a market based social security system most Latin
American countries in the nineties would introduce mandatory individual capitalization funds.
There is though and important lesson to be learned from Chile as it advanced towards democracy.
The “Concertación” government despite their opposition to Pinochet did not want to change a model that
seemed to be working on the economic side, yet they recognized the importance of addressing the social
question after years of neglect. And they needed resources for that. The engineering of a pact between the
political elites and the economic elites after the transition to democracy, in order to increase taxes to
finance expansionary social policies remains an example that with few other suggest the mutually
42
reinforcing character of social policy investment and democracy. Once again a liberal stage and the
disarticulation of historical corporation paved the way to a broad ranging agreement regarding taxes and
social protection.
b. The 1990s: A peculiar liberal route of reform: social spending and social security.
i. Recovering social spending
If the 1980s were characterized by low spending, absence of structural social state reform and
emergency funds strategies, the 1990s would imply a definite structural shift with an articulated liberal
agenda. On the other hand the 1990s were also times of expanding social spending. On average for 17
Latin American countries, public social spending went from 350 dollars percápita to 550 dollars
percapita.
The analyses performed by ECLAC on the evolution of social spending shows that the reasons
behind this important increase are not due solely to GDP growth or to an overall expansion of state
spending. A large part of the new dollars for social policy came from an increased fiscal priority of social
policies. What may even be more important, after 1997 when growth slowed down and so did total state
expenditures, social spending continued to crawl upwards until 1999. The deepening of the regional
economic crises brought asocial spending to a halt in some countries, but remained moving slowly
upwards in others. All in all as a percentage of GDP the 1990s showed a robust recovery of social
spending. This also translated into better percapita expenditures, even in amplified form, since 1990-1997
was a period of GDP growth in Latin America.
TABLE 9
Social expenditure variations as a percentage of gdp for latin american countries.
Argentina
Brasil
Chile
Colombia
Costa Rica
Honduras
Mexico
Nicaragua
Panama
Paraguay
Peru
1990-1992 1992-1994 1994-1997 1997-1999
8.0%
9.0%
-6.0%
4.0%
-2.0%
13.0%
-2.0%
7.0%
5.0%
0.0%
6.0%
11.0%
18.0%
22.0%
33.0%
-2.0%
-3.0%
5.0%
6.0%
-1.0%
8.0%
-9.0%
-6.0%
3.0%
25.0%
9.0%
-3.0%
7.0%
-2.0%
19.0%
-13.0%
15.0%
5.0%
2.0%
6.0%
-7.0%
10.0%
13.0%
14.0%
-7.0%
45.0%
21.0%
5.0%
11.0%
43
Uruguay
Venezuela
Non weighted average
12.0%
-1.0%
16.8%
7.0%
-15.0%
7.4%
3.0%
9.0%
4.0%
9.0%
4.0%
4.2%
Source: CEPAL, Panorama Social de América Latina, 2000-2001.
If democracy had been unable in the eighties to stop the fall of social spending, in the nineties,
partially helped by the economy, social spending recovered vigorously, though towards the end of the
decade already showed signs of slowing down. A large part of this increases went to health care and
education, allowing for an important expansion of these services. Yet for reasons of space and time I will
not concentrate on education and health care reforms. They are indeed relevant but they require,
especially education, a paper on and of itself. Health care will be considered but as a close sibling of
social security reform. In many countries the share of the lion in this rapid increase in social spending
went precisely there, and did so, not just to finance old obligations, but also to cover the costs of
structural transformations. To those transformations and the interplay of democracy and reform we now
turn.
ii. Relevant nuances and their political keys in the liberal trajectories in social security.
The 1990s placed Latin American as a definite world laboratory on pension reform. After the
pioneering reform of 1981 in Chile, little had happened for the rest of the decade besides some relevant
parametric reforms in Uruguay and Argentina. Yet in the 1990s the liberal agenda came with a vengeance
in the region. If we count Nicaragua and the Dominican Republic who have legislated though not yet
implemented reform, eleven countries introduced compulsory individual capitalization funds administered
by private firms or by private and parastate entities11. There are three central new and powerful actors to
be considered when we analyze the causes behind social security reform in Latin America in the late
eighties and throughout the nineties: local and international financial interests, multilateral agencies, and
democratically appointed techno-political elites with a strong representation of economists. In addition
there are traditional actors that operate and interact with these new players: trade unions and beneficiaries
from the old programs of the social state. And there is a stage where these players made their moves: with
the exception of Chile (and to a certain extent Peru) major reforms were carried out in democratizing or
democratic and fiscally constrained yet not fiscally bankrupt contexts and with credit lines for certain
types of reforms and social policy innovations. Such stage and such players led to a peculiar mix of
11
Chile did it as we said in 1981, Peru in 1993, Argentina in 1994, Uruguay in 1996, Mexico, El Salvador and
Bolivia in 1997 and Costa Rica in 2001.
44
reform. A play where on the one hand previous privileges were partly maintained, while future
beneficiaries entered a more market-based, rather than occupationally and status-based social security
systems. Privatization along the lines of liberal regimes combined with the persistence of bismarckian and
many times dualist models were sought as the predominant routes of reform. In the cases were structural
reforms did not take place two major strategies were considered with dissimilar results depending on the
countries: parametric reforms and homogenization reforms.
Among the countries that sought liberal lines of reform we find countries that belong to the three
different types of social states that had developed between the 1930s and 1970s. Indeed what is most
surprising in these trends of reform is that countries facing widely different challenges in their social
protection systems leaned toward the private solution (Mesa Lago, 2002, 2004; Muller, 2003). Yet it is
also true, that with the exception of Chile, the two other universalistic stratified countries and also Costa
Rica, all under a democratic regime assumed the mildest form of privatization of social security. Among
the dualist countries only Mexico moved to a private system and did so much more radically than the
universalistic countries, closing down the old system, and did it still under the PRI administration. Brazil
under a democracy and with the constitutionalist movement of 1988 did not seriously consider at any
point privatization. Peru under the “constitutional dictatorship” Fujimori, an with a legacy of peculiar
truncated case of dualism, moved to the private system but maintained the old system, with minor
parametric reforms as a competitor. Yet by increasing contributions in the old public tier, Peru provided
central incentives for people to move to the private option. Furthermore, all new contributors enter
directly into the private system. In short the public system today is being closed down with less than 5%
of all contributors. Finally those exclusionary countries that moved towards privatized social security
systems, Bolivia and El Salvador, did so radically, closing down or phasing out the previous pay as you
go systems.
Yet this transformation of social security systems was less liberal that one could assume at first
sight. The main reason is that most countries did not include in these reforms some special privileged
groups from the past. The following table based on Mesa Lago´s thorough and keen evaluation of the
reforms combines information of the shift to liberal systems in Latin America and the remaining legacies
that each system kept from the times of privilege.
A second reason is that many countries did not incorporate what is today a standard instrument of
liberal regimes for dealing with those that do not reach the qualifications to be granted a pension: a noncontributory means tested social assistance pension. Some of the countries that adopted liberal models do
offer a state guarantee that if the contributor does not reach an adequate contribution and thus pension
value, the state will provide the difference up to a basic pension. Yet this is different than the social
45
assistance non-contributory pensions, since for eligibility some years of work and contributions are still
required. No democracy, with the exception of Costa Rica and Uruguay, dealt with the issue of noncontributory protection as a major issue of the reform, yet all the democratizing or democratic regimes
honored old time beneficiaries and lobbies leaving them out of the reform or buying with additional
privileges their non-opposition.
TABLE 10
PAYG Tier
Uruguay
Remains
Argentina
Remains
Costa Rica
Chile
Remains
Closed
down
Closed
down
Competes
Mexico
Role
of Remaining privileged systems
private tier
Second tier
Armed
Forces,
Bank
employees,
Notaries, State University.
Second tier
Elected officials, judiciary, diplomats,
provincial civil servants, independent
professionals.
Second tier
Judiciary and Teachers
Substitutive
Armed Forces
Yes
Yes
Substitutive*
No
Civil servants (federal, state and
municipal), oil company workers.
Colombia
Competes
Most of the Public Sector, elected
officials, oil company workers, teachers.
Peru
Competes
Competes
State employees (phasing out)
El Salvador
Phased out
Substitutive
Armed Forces
Bolivia
Closed
Substitutive
Armed Forces go to new system but with
down
privileges
Sources: Mesa Lago, 1999; Mesa Lago 2004, Muller, 2003
Social
Assistance
Yes
Yes
No
No
No
No
* Protection for Sickness and Death was not privatized.
The first questions that naturally comes to mind is: why didn’t the liberal reform hailed as a better
deal for future beneficiaries include those who control, run and or have the coercive means of control of
the state? The question is of course rhetorical. Under the liberal regimes they would loose part or all of
their previous privileges. These privileges include in most cases better replacement rates, earlier ages of
retirement and broader coverage of personal and family risks that are covered by state resources beyond
contributions or by private state endowed tariffs for services (independent professionals).
Another question that is relevant to pose and that might give us some clue as to what were the
politics of social security reform is why under such different previous regimes did all countries
incorporate, albeit, to differing degrees private capitalization funds. Part of the answer rests on the major
political force behind these processes of privatization: the multilateral lending agencies.
46
It is indeed true that the universalistic countries shared a common problem of chronic deficit and an
active population to pensioners’ population ratio that made it hard to sustain a pay as you go system into
the future. It is also true that this situation had led the universalistic countries to finance social security
only partly from contributions. In Argentina and Uruguay previous to the reform almost half of all social
security revenues came from general taxes, not from contributions. Technically this allowed for three
basic solutions: individual capitalization funds, collective capitalization funds or a shift to a general tax
based financed social security system. Both the second and third choice allows for intragenerational
distribution of rent and risks, the former does not (Huber, 1996). The choice was always for the former.
Only the universalistic countries with the exception of Chile allowed for a redistributional dimension in
the form of PAYG systems. Besides the MLAs, the reasons for such a choice lays additionally in three
major players: domestic and international financial capital, old privileged groups in the previous social
security system and neoclassic trained economist in the Ministries of Finance, Central Banks and
Economics.
The influence of the Chilean example and of the World Bank publication “Averting the Old Age
Crises” cannot be sufficiently stated. But there were also concrete pressures from the MLA for countries
to push this type of reform. For political elites this allowed for an easy way out of their social security
crises. Loans would help cover the costs of transition, which is simply another way to say, that they
would subsidized the phasing out of the old systems, while the promise of better pensions could be made
with a time frame for contrasting such claim well ahead of their periods in office (Kay, 1999). In
Argentina and Uruguay the increasing organization and actions of the retired people’s movement placed
mounting fiscal pressure on an already bankrupt system. The defense of the old system by these
constituencies tended to struggle for better pensions within the old stratified social security system. The
left usually aligned themselves with these movements, as did the trade unions. Government on the other
hand and their political parties increasingly looked at MLA´s both for financial support and technical
support.
While in most countries either the crises of social security or fiscal and economic crises helped set
the stage for reform, the driving force behind such reforms came from domestic economic elites and
international financial agencies.
The Bank Association of Argentina had made a proposal for privatization as early as 1992, and the
IADB and the World Bank had also explicitly pushed for privatization Chilean style. In the agreement of
1992 with the IMF it was explicitly stated the need and commitment of the Argentinean government for
privatization (Muller, 2003). In Uruguay a consultancy from the IADB drafted in 1993 a document
indicating that parametric reforms would not be sufficient to deal with the crises of social security.
47
Increasingly the political arena in Uruguay became divided with one side composed of the left, the trade
union movement and the retired people’s movement defending the old system, and the governments
trying to find a way out of the crises (Filgueira y Moraes, 1999). In 1995 a team of experts that came with
3.8 million dollars loan promise from the IADB and a team of government experts worked for three
month, isolated from the political process, coming out with the proposal that would finally be approved.
Democracy helped moderate and at the same time keep the privileges of old beneficiaries in the
systems of Argentina and Uruguay. The case of Costa Rica shows a different pattern. Thought as a
needed reform and also with definite assistance of the MLAs, Costa Rica also developed a wide-ranging
negotiation with social and political actors at the same time that it pushed for a reform that would increase
coverage. The end result is a liberal reform with egalitarian undertones (Martinez and Mesa Lago, 2003).
Of the 19 privileged systems for pensions of the past only two remain: teachers and the judiciary, and the
non-contributory pensions remain as a major pillar of the system (Mesa Lago, 2004).
In the case of Peru, the need to gain the recognition of the MLAs made Fujimori a staunch defender
of structural reforms, where social security was one of them. Once again the economists, the MLAs and
the political need of the executive guaranteed a rapid reform, the first to be carried out in Latin America
after the Chilean reform. After the coup of Fujimori in 1992 the stage was set for reform. Yet after
publishing the draft of the reform law serious opposition from the beneficiaries of the old models and
from the trade unions limited the reform proposal and ended up maintaining the old system parallel to the
new one. Yet a number of incentives implied that by 2001 the public system was bankrupt and small. As
of today Peru is steadily moving towards a purely private system (Muller, 2003). Bolivia’s path to reform
was marked by the open opposition of trade unions and left wing parties and by the support of the
business and financial community. The program of reform was also linked to a broader privatization plan
(capitalization plan in the words of the reformers) and once again relayed on the financial and technical
support of the world bank and the technical support of Chilean experts. Once again this reform met with
the opposition of trade unions and retired people. In its Capitalization Program Adjustment Credit the
World Bank made future loans conditional on pension structural reform. In 1997 the law was passed and
despite constitutional complaints by beneficiaries has so long stayed in place (Muller, 2003).
In Mexico the crises of 1994 open the road for wide ranging attempts at structural reform. The idea
behind pension privatization was not simply to deal with a system that was approaching fiscal problems,
but more to open opportunities for investment and capital markets. The proposal by the Zedillo
government of 1994 did not face the opposition of the workers confederation, but did have to confront
both opposition parties and the very vocal trade union of the Mexican Institute of Social Security. Yet the
support of the entrepreneurs that saw their contributions diminished and the support of the MLA´s tilted
48
the balance and allowed for the law to be approved. Besides capitalization this was an important
parametric reform that made eligibility much more harder by drastically increasing the needed year of
contribution. Finally, it must also be noted that in the case of Mexico, together with this privatization
reforms, the old emergency fund strategies of the 80s persisted and enhanced their aims and population
coverage. Pronasol, would turn into Progresa and then into the Fox government program Oportunidades.
This program targets the poor (today only the rural and semi rural poor) and transfers rents in exchange
for school attendance and health check ups for mothers and children in this population. At present it is
one of the most important social programs in the country with coverage of 4.6 million poor households.
Once again despite the dual characteristic of the Mexican social state, the space for rural pro-poor policies
remains.
Given this very scant description of the politics and major features of social security reform in the
countries that introduced liberal reforms, some patterns are worth mentioning.
§
All reforms counted with the technical and financial support of MLAs. In some conditionality
was a major instrument pushing reforms.
§
All countries faced fiscal or economic crises, some located within the social security system,
some beyond it.
§
In all countries the business community was either an open supporter or a sympathizer of the
reforms. Consistently in most countries entrepreneur’s contributions either disappeared or were
clearly diminished as a basis for financing social security.
§
With the exception of Mexico, were there is not a clear opposition from trade unions, labor
opposed these reforms, as did old beneficiaries. Most beneficiaries from the old models were
able to see their entitlements honored and in some cases kept their old systems squarely in
place.
§
No alternative proposal to the liberal model was either made or gained adequate visibility. The
left, trade unions and beneficiaries, defended the old model and pressured for increased
resources, rather than came up with alternatives.
§
The non-contributory social assistance pensions were rarely a major issue in the reform
proposals. The case of Bolivia with Bonosol and the case of Costa Rica might be the only
exceptions. Clearly no party and no social movement pushed forward or at least were able to
place that area of reform in the center of the agenda.
§
Indeed, the mottos of reform were about fiscal, economic, efficiency and equity concerns
(understood as getting out what you put in), not about coverage and equality.
49
Democracy and politics played a definite role in these reforms, but did so in a rather perverse
fashion. The old state constituencies of the ISI model were usually able to hang on to a large part of their
privileges especially where they were either a part of the powerful old lobbies or a large electoral
constituency, while private workers were moved to the capitalization funds. The absence of representative
parties and of ideologies that looked at social policy as a matter of social inclusion rather than poverty
alleviation, left the historically and newly excluded sectors without institutionalized forms of protection
or very weak forms of protection. On the other hand the need of democratic governments to balance their
budgets and have a positive credit line with MLA´s led them to support structural adjustment packages
that included the newly privatized creed on social security. By shifting from PAYG system to private
capitalization systems and by getting financial backup in the form of credit to finance the old deficit prone
system, political elites bought time well beyond their times in office (Kay, 2002). Both fiscal costs of
present days reforms will only be known in the future when countries have to pay their debts with MLAs
and when the new capitalization private system start to pay their pensions. These costs have indeed been
pushed well into the future.
Regarding social costs, most models built to predict the effect of these reforms on future social
situations of the elderly suggest that both coverage and inequality will suffer. A simple table on the
percentage of population covered by one or two tiered models clearly points to the deficit in coverage that
these systems will have to face in the future. True, much of the poor performance in terms of coverage
this systems confront are not of their own making but rather depend on the interactions of their eligibility
criteria and the development of the labor market and of formal employment in the last decade and a half.
TABLE 11
Coverage before and after structural social security reforms measured by percentage of the population
making contributions to social security
Coverage by actual Affiliates considering Coverage
by
actual
contribution
contributions
all tiers after reform
considering all tiers after reform
before reform
(2002)
(2002)
Chile
64
111
58
Argentina
50
69
24
Uruguay
73
77
60
Costa Rica
53
65
48
Mexico
37
72
30
Colombia
32
59
24
50
Bolivia
12
23
11
El Salvador
26
40
19
Perçu
31
28
11
Source: Mesa Lago, 2004
A question remains: will the state honor its pledge to guarantee a minimum pension and
eventually extend such pledge in the form of a flat rate non contributory social assistance pension for all
those that seem to be in the future non protected population. The fact that today most of these regimes are
democracies, and if they remain so for the next years, suggest that this might well be a possibility, since
almost 50% in most of these countries will reach old age without adequate social security protection. If
this happens, then the road open for a more egalitarian and inclusionary social security system might be in
the wings, even though its fiscal cost might be unbearable for many of these countries. The cases of
Uruguay, Costa Rica and to a lesser extent Argentina are best positioned to attempt such shift. They kept
their PAYG systems and they all have some form of social assistance non-contributory pensions. If such
becomes the first integrated pillar with collective capitalization funds or with flat rate benefits (be them
means tested or universal) financed out of general taxes then basic universal income protection might be
closer than what the last twenty years of reform might lead us to expect. Oddly, Chile, the most liberal of
all regimes might also be moving in that direction. The fact that only 50% of the population is currently
contributing to the private system leaves and immense amount of people with no protection in their old
age. In addition Chile, Brazil, Uruguay, Argentina and Costa Rica also have some form of rent guarantee
for the working population with children (Mexicos´s Oportunidades also has developed such a system)
and in the case of Argentina they have also developed a sui-generis form of unemployment insurance with
the plan “Jefes y Jefas”.
The case of Brazil: averting the private utopia crash and in the road from dual to basic
inclusionary universalism?
The Brazilian dictatorship of 1964 inherited a stratified dual regime based on the ISI model. In contrast
with the later authoritarian regimes of the southern cone, the Brazilian government did not pursue neoliberal reforms, but rather deepened the ISI model. In doing so, and doing so with a certain degree of
success, Brazil continued to modernize its social structure, and with it diminish its dual, though by not
means, stratified nature. Yet by 1970 almost thirty percent of the urban population and most of the rural
population had no form of protection in the Brazilian system. But the military regime, besides pushing the
51
ISI agenda, did two things that would start to redefine the spirit if not the reality of its social state. First
between 1966 and 1976 the state unified and standardized a good part of the social security system under
the authority of the Ministry of Social Insurance and Social Assistance. While this did not imply a
suppression of benefits it did take the old corporations out of the management of the system. Secondly it
pushed for the expansion of coverage of the non-organized population, especially the rural worker, by
creating the rural pensions fund (FUNRURAL) with no contributory base on the side of the rural worker
and by also expanding health care to the previously uninsured urban population and part of the rural
sectors (Fleury, 2001). Despite the fact that resources were not enough to turn these initiatives and
translate them immediately into major transformations in terms of coverage, they generated three positive
developments: they set the stage for the democratic governments to come, they fueled the health
movement (sanitarista movement) with doctors and medical personnel that entered as civil servants into
these new social protection systems, and they also contributed, later, even in the authoritarian regime to
an important expansion in coverage of the rural population contributing to the process of politization and
mobilization of the rural base (Dos Santos, 1979). Maybe the most important aspect of FUNRURAL is
that it veers away from the notion of social citizenship based on formal wage citizenship. In doing so,
Brazil had made a major leap towards the possibility of a basic inclusionary universalistic system.
The democratic transition of Brazil brought with it an effervescence of participation and
organizations, some from the old ISI model, but many from the new constituencies and the historically
excluded. The constitution of 1988 would for the first time recognize social rights as inherent citizenship
rights, and such result is indeed the direct product of a broad alliance between opposition parties to the
regime and new social movements that emerged with democratization. Yet together with it came new
democratic pressures from the powerful old beneficiaries of the social security system. While the social
security law defined three universal pillars based on health, social insurance and social assistance, a large
part of the social protection system was still going to cover the 1.000.000 plus retirees of the privileged
state structures in an altogether differentiated system. Also the very letter of the law recognized two
different principles, one of citizenship and one of insurance. Most of the income transfers remained in the
area of social security while social assistance would grow but at a moderate rate. It was in the area of
health care that the most radical transformation took place, creating the Unified Health System (Sistema
Unico de Saude) granting free, universal access to all the Brazilian population. The sanitarista movement
started in academia and with a part of the medical corporation became the leaders of a wide political
movement pushing to turn the letter of the constitution into policy and resources.
Yet Brazil tends to write its laws ahead of its resource base and resource commitments, and thus
real access and quality many times suffers the consequences. The popular phrase “paper citizenship” that
52
many Brazilians use, illustrates the point. Still, coming from a dual and stratified system, Brazil has made
important progress towards a more egalitarian basic inclusionary system (Draibe, 2002). The passing of
the Organic Law of Social Assistance and the creation and recent unification under the name “bolsa
familia” of a pension for poor families that joins the old age social assistance pensions are also steps in
the same direction. The recent reform of social security, limiting and in some cases suppressing the
privileges of the old state insurance systems advances towards the same goals.
Only after the old corporations that were in the base of the ISI model of social protection had
been attacked by the military, and only after a new set of parties, one with technocratic social democratic
leanings oriented politicians (PMDB) and another that combined constituencies from the old workers of
the ISI model and the historically excluded (PT) did the politics of social protection shift from narrow
corporatism, not to liberal privatization, but to embryonic basic inclusionary universalism.
IN CLOSING
THE TURN OF THE CENTURY: REALITY AND PROSPECTS OF SOCIAL POLICY AND
DEMOCRACY IN LATIN AMERICA
Universalistic, dualist and exclusionary countries no longer fit their 1970s molds. New models
have emerged in Latin America. Chile stands alone as a universalistic liberal residual regime that has
attempted steps towards a less inegalitarian social protection system. Costa Rica and Brazil seem to be
making strides towards a basic protection inclusionary regime with stratification at the top, but are still a
long way to go, given previous privileges and the scarcity of quality broad coverage in non-contributory
social services and social pensions. Uruguay and Argentina after their liberal turn, and because they
retained both the PAYG systems and the social assistance pensions might also move in that direction.
Mexico stands at the crossroads with a model that seems to have two purposes: efficiency and
fiscal responsibility with limited coverage for present population, and targeted integrated programs for the
poor, aiming more at breaking the intergenerational transfer of poverty than at combating poverty in
present days (Brachet, 2004). Yet in the case of Mexico, Argentina and Uruguay there is also the need to
address the issue of special privileges from the old systems.
As for the exclusionary countries the expansion of social expenditure has helped them increase
coverage, especially in what relates to education and health, yet regarding their social security systems,
little if any advances have been made. Those that chose or are choosing the liberal privatized model
choose a system that even in more advanced countries has been unable to deal with the problems of
53
coverage. A flat rate general tax based means tested or universal benefit remains for most of Latin
America the best promise of an inclusionary social state.
Both the practical, technical and institutional side of such an options are available after the
downsizing and destruction of the old corporatist models. The political will, though will not only be
needed at the domestic level, with parties that address broad interests rather than narrow lobbies, but also
at the transnational level, requiring a more drastic humbling of the MLAs as to their responsibility and
future action regarding twenty years of misguided developmental turns in regards to the social state.
Neither social emergency funds, nor fiscally restrictive budgets and even less private and market oriented
solutions to social protection have proven adequate for the immense challenge of incorporation that lays
ahead for the Latin American states and people.
54
Bibliography
Avelino, George (1997); Democratización y Gasto Social en América Latina, 1980s-1990s; en Andrés
Perez Baltodano (ed.) Globalización, Ciudadanía y política social en América Latina: tensiones y
contradicciones. Nueva Sociedad: Caracas, pp. 97-122..
Brachet-Márquez, Viviane (2004); Nacimiento, auge y transformación del Estado benefactor mexicano
(1823-2000), UNRISD Development Project, Social Policy in a Development Context
Bulmer Thomas, Victor (1996): “Introduction” en Bulmer Thomas ed.; The New Economic Model in
Latin America and its Impacto on Income Distribution and Poverty, McMillan Press, Londres.
CEPAL (1994): Panorama Social de América Latina. CEPAL, Santiago de Chile.
CEPAL (2000-2001): Panorama Social de América Latina. CEPAL, Santiago de Chile.
CEPAL (2001-2002): Panorama Social de América Latina. CEPAL, Santiago de Chile.
Cominetti, Rossella (1994); “Gasto Social y Ajuste Fiscal en América Latina” en Serie Reformas de
Política Pública, 20. Santiago de Chile: CEPAL.
Deacon, Bob (1994); “Global Institutions, Social Policy and Social Development”. Working paper para el
Summit de las Naciones Unidas sobre Desarrollo Social. Leeds, UK; United Nations Research Institute
for Social Development.
Draibe, Sonia (2004); The Brazilian Developmental Welfare State: Rise Decline and Perspectives.
UNRISD Development Project, Social Policy in a Development Context.
Draibe, Sonia M. (2002), “Social policies in the Nineties”. In: BAUMANN, Renato (ed.), Brazil in the
1990s – An economy in transition. Houndmills, UK, Palgrave Publishers (St. Antony´s Series).
Draibe, S. M. (1985); Rumos e Metamorfoses. Um estudo sobre a constituição do Estado e as alternativas
da Industrialização no Brasil 1930-1960, Ed. Paz e Terra, São Paulo.
Esping Anderson, Gosta (1990); The Three Worlds of Welfare Capitalism. Princeton: Princeton
University Press.
------------------------------ (1994); “Welfare States and the Economy” en Smelser & Swedberg (ed.) The
Handbook of Economic Sociology. Princeton University Press; Princeton.
Filgueira, Fernando (2002): “Latin America: Society and Comparative Studies” en International
Encyclopedia of Social and Behavioral Sciences, pp..8430-8436. SSRC, UK.
------------------ (1998): “El nuevo modelo de prestaciones sociales en América Latina: residualismo,
efciencia y ciudadanía estratificada” en Brian Roberts (ed.) Ciudadanía y Política Sociales,.
FLACSO/SSRC: San José de Costa Rica
------------------ (1997): “La nueva arena de las políticas sociales: vectores internacionales y mediación
doméstica en la reforma del sector social en América Latina” en Andrés Perez Baltodano (ed.)
55
Globalización, Ciudadanía y política social en América Latina: tensiones y contradicciones. Nueva
Sociedad: Caracas.
------------------ y Martinez, J. (2002): “Paradigmas Globales y Filtros Domésticos: Reforma
Administrativa y Política Social en América Latina” en Ciencias Económicas, Vol. XVIII, Nro. 2, pp.
191-215, San José de Costa Rica.
------------------- y Filgueira, C. (2002): “Models of Welfare and Models of Capitalism: the limits of
transferability” en Evelyne Huber (ed.); Welfare Regimes and State Reform in Developing Countries”;
Penn University Press,.
-------------------- y Moraes, J.A. (2000): “The Political Economy of Institutional Reform in Uruguay:
Enabling Conditions, Strategic Devices and Sector Specific Configurations”. Working Paper Series. BID,
Washington.
------------------- y Papadópulos, J. (1997): "Putting conservatism to good use? Neoliberal Transformations
in Uruguay" en Douglas A. Chalmers, Carlos M. Vilas, Katherine Hite, Scott B. Martin, Kerianne Piester
y Monique Segarra (eds.).;The New Politics of Inequality in Latin America. Rethinking Representation in
Latin America, Oxford University Press; Oxford.
Fitzgerald, E. V. K(1996): “The New Trade Regime. Macroeconomic Behavior and Income Distribution
in Latin America” en Bulmer-Thomas (ed.) op. cit.
Fleury, Sonia (1994); Estado sem Cidadaos Seguridade Social na America Latina. Editora Fiocruz; Rio
de Janeiro.
------------------(1999); Reforma del Estado en América Latina. Hacia donde?. Nueva Sociedad, Caracas,
Nro. 160, 58-80.
Fleury, Sonia (2001), "Universal, Dual or Plural? Health Care Models and Issues In Latin America". IN:
Molina, Carlos Gerardo & José Núñez del Arco org. Health Services in Latin America And Asia.
Washington D.C.
Fleury, S. y Molina C.G. (2000). Modelos de Protección Social. INDES. Manuscrito. Washington, D.C.:
Banco Interamericano de Desarrollo: Instituto Interamericano para el Desarrollo Social.
Huber, Evelyne (1994); “Social Policy in Latin America: Welfare State Aspirations, Crsis and Prospects”.
Mimeo: University of North Carolina.
--------------------(1996); “Options for Social Policy in Latin America: Neoliberal vs Social Democratic
Models” in Esping Andersen (ed.)Welfare States in Transition. National Adaptations in Global
Economics. UNRISD. Sage; London.
Kay, Stephen (1999); The Politics of Postponement: Political Incentives and the Sequencing of Social
Security Reforms in Argentina and Uruguay. Research department. Federal Research Bank of Atlanta.
Korpi, Walter (1983); The Democratic Class Struggle. Routledge and Kegan Paul; London.
56
Malloy, James (1985); “Statecraft and Social Security Policy and Crisis: A Comparison of Latin America
and the United States” in Mesa Lago (ed.) The crisis of Social Security and Health Care. university of
Pittsburgh. Pittsburgh.
Mesa Lago, Carmelo (1978); Social Security in Latin America: Pressure Groups, Stratification and
Inequality. Pittsburgh: University of Pittsburgh Press.
Mesa-Lago, Carmelo (1991); “Social Security in Latin America”. Informe elaborado para Inter-American
Development Bank, Economic and Social Progress in Latin America, 1991, Report.
----------------------------- (1994); Changing Social Security in Latin America. Toward Alleviating the
Costs of Economic Reform. Boulder and London: Lyenne Rienner Publisher.
------------------------------ (1999); Old-Age Security and Health Care for the Poor in Latin America and
the Caribbean. IADB, Conference on Social Protection and Poverty.
------------------------------- (2004); Las reformas de pensiones en América Latina y su impacto en los
principios de la seguridad social. CEPAL, Serie Financiamiento del desarrollo # 144.
Muller, Katharina (2003); Privatizing Old Age Security. Latin America and Eastern Europe Compared.
Edwarg Elgar, Cheltelham.
Miranda R., Ernesto (1994); “Cobertura, Eficiencia y Equidad en el Area de Salud en América Latina.
Problemas y Propuestas de Solución” en Serie Políticas Sociales, Nr. 5. CEPAL; Santiago de Chile.
Przeworski et all (2000): Democracy and Development. Political Institutions and Well-Being in the
World, 1950-1990. CAmbridge University Press : Cambridge, London, New York.
Rueschemeyer, Stephens & Stephens (1992); Capitalist Development & Democracy. University of
Chicago Press; Chicago.
Sagasti, F; Prada, F and Bazán, M. (2004); Social policy in a Development Context – Peru Report.
UNRISD Development Project, Social Policy in a Development Cont
Santos, Wanderley G. dos. Cidadania e justiça. Rio de Janeiro. Editora Campus.1979
Titmuss, Richard (1958); Essays on the Welfare State. Allen and Unwin; London.
Williamson, John (ed.) (1990); Latin American Adjustment: How much has happened? Washington D.C.:
Institute for International Economics.
World Bank (1994); World Development Report, 1994. Washington, D.C.
Wurgaft, José (1993); Fondos de Inversión Social: América Latina. Santiago de Chile: PREALC/OIT.
57
58